Carnosine

A small natural molecule made of two building blocks joined together — beta-alanine and histidine — found in meat and in your own muscle and brain, and sold as a food supplement in capsules for ageing, exercise and eye health. The sequence field is deliberately left empty: beta-alanine is not one of the twenty standard amino acids and has no letter in the one-letter code, so any string written here would describe a different molecule. PubChem gives C9H14N4O3 and 226.23 g/mol (CID 439224, CAS 305-84-0). Two related products are sold under near-identical names and are different substances: N-acetylcarnosine eye drops, and zinc-carnosine (polaprezinc), a stomach medicine approved in Japan since 1994.

Sold as

Swallowed, as a capsule or powder. Sold separately as N-acetylcarnosine eye drops and as zinc-carnosine for the stomach — both different products.

Legal status in the EU

Not approved as a medicine in the EU or the United States: the US regulator's approved-drug and drug-label databases return no product with carnosine as an active ingredient, and this review found no European marketing authorisation. It is sold as a food supplement, which means no regulator checks that it works before sale. Zinc-carnosine, generic name polaprezinc, is a different substance and has been an approved anti-ulcer medicine in Japan since 1994; that approval says nothing about carnosine itself. Neither carnosine nor beta-alanine appears anywhere in the World Anti-Doping Agency's 2026 Prohibited List, which was searched word by word on 4 August 2026.

In muscle, where the body keeps most of it, carnosine works as a buffer that soaks up the acid produced during hard, short efforts, and it also mops up reactive molecules and blocks sugar from sticking to proteins. The problem with swallowing it is an enzyme in human blood called carnosinase, which cuts it in half within minutes; how much of that enzyme a person has is largely inherited, and in one study only 8 of 25 healthy adults given a large dose showed any measurable rise in blood carnosine at all. This matters for how the compound is sold: the beta-alanine half released by that breakdown is what muscle takes up and rebuilds carnosine from, which is why raising muscle carnosine is done with beta-alanine rather than with carnosine.

Part of the bodyWhat happensHow it is knownKind of evidence

Blood, in the hour after you swallow it

Carnosine does reach the blood — that part is settled. What happens next is the whole problem. Human blood carries an enzyme called carnosinase that cuts carnosine in half, into beta-alanine and histidine, and its working life in blood is measured in minutes rather than hours. Blood levels peak within about an hour and there is very little left after four. How fast this happens differs a lot between people, and the difference is largely inherited: in one study only 8 of 25 healthy adults given a large single dose showed any measurable rise in blood carnosine at all, and those who did had roughly half as much of the enzyme as those who did not.

A dose-escalation study in 16 healthy volunteers given single doses of 4, 6, 10 and 15 g with two-week gaps, tracking blood levels — a 20 g step was written into the protocol but never given, because the escalation was stopped at 15 g; a study in 25 healthy adults given 60 mg per kilogram of body weight in one go; and an older feeding study in 18 adults given 200 g of cooked minced beef containing 124 mg of carnosine per 100 g, in whom blood carnosine peaked at 32.7 mg per litre 2.5 hours after the meal and was undetectable by 5.5 hours.

Measured in people

Source [03]Source [04]Source [17]Source [18]

Muscle

Your muscles already hold carnosine, and how much varies about fourfold between people. It works there as a buffer: it soaks up the acid that builds up during hard, short efforts. Swallowing carnosine does raise the amount stored in muscle — but not by arriving there intact. It is cut apart in the blood first, and the beta-alanine half is what the muscle picks up and rebuilds carnosine from. That is why the supplement industry sells beta-alanine rather than carnosine for this purpose: you can give far less of it for the same effect. Which muscle carnosine you start with is mostly about who you are, not what you take: men measured 28% to 82% higher than women depending on the muscle, the level falls with age, and vegetarians measured 26% lower than meat-eaters in the calf.

A four-week supplementation study in which the group swallowing L-carnosine — at an amount containing the same quantity of beta-alanine as the 6.4 g a day beta-alanine group — ended with muscle carnosine 65.8% higher than where they started, against 64.2% in the beta-alanine group. The starting-point comparison comes from measurements in 149 healthy people, 94 men including 12 vegetarians and 55 women, with muscle carnosine read by magnetic resonance scanning of three leg muscles.

Measured in people

Source [05]Source [19]Source [20]

Blood sugar

This is where carnosine itself, swallowed, has the most consistent controlled evidence. In adults with prediabetes or early type 2 diabetes, blood sugar measured 90 and 120 minutes after a standard sugary drink came down further on carnosine than on a matching dummy capsule, and the total sugar load over the two hours was lower. Insulin did not change, which is why the authors think the effect is on how much sugar the liver puts out rather than on the pancreas. In a separate trial in people already diagnosed with type 2 diabetes, fasting sugar and the three-month blood sugar average both came down further than on the dummy.

Two randomised double-blind trials against a dummy capsule: 43 adults with prediabetes or type 2 diabetes on 2 g a day for 14 weeks, and 54 adults with type 2 diabetes on 1 g a day for 12 weeks. An earlier pilot in 30 adults with overweight or obesity and no diabetes, on 2 g a day for 12 weeks, pointed the same way. All three are small, none ran longer than 14 weeks, and none measured whether anyone got ill less often.

Measured in people

Source [01]Source [11]Source [21]

Thinking and memory

Carnosine sits in the brain as well as in muscle, and it has been tested as an add-on to normal psychiatric treatment. The results do not line up. In a 75-person trial in schizophrenia, two of a battery of tests came out better than on a dummy capsule and the rest showed no difference. In a larger 100-person trial designed around a different question — whether it improves the withdrawn, flattened side of the illness — there was no difference from a dummy at all, and only an attention score separated at the higher dose. A 2026 review that gathered every carnosine trial it could find reported that five trials of carnosine on its own found no difference from a dummy on most of the thinking tests they used.

Two randomised double-blind trials against a dummy capsule: 75 adults with stable long-term schizophrenia on 2 g a day for 3 months, and 100 patients on 400 mg a day for 12 weeks then 800 mg a day to week 24. Pooled alongside them, a 2026 review of 13 studies in which 6 used carnosine alone, 6 used it mixed with anserine or antioxidants and 1 used the zinc form. That review rated the strength of the evidence low or very low throughout, and it was paid for by a company that sells a beta-alanine ingredient.

Studies in people disagree

Source [10]Source [15]Source [22]

Children with autism

This is the use carnosine became known for, and it is the clearest example on this page of an early claim not holding up. A 2002 trial of 31 children reported that the children given carnosine improved on a standard autism rating scale over eight weeks while the children given a dummy did not. That is a comparison of each group with its own starting point, not a comparison of the two groups with each other. When five later trials in 215 children were pooled and the groups were compared directly, the difference between carnosine and a dummy was not distinguishable from chance on either of the two rating scales used.

Pooled from 5 trials in 215 participants, 4 of them double-blind and placebo-controlled. Individual trials used 500 mg to 800 mg a day for 8 weeks to 2 months. The reviewers' own conclusion is that current data do not support using carnosine in children with autism, and that this is partly because there are so few trials and they are so small.

Studies in people disagree

Source [02]Source [23]Source [24]

Kidneys, in children with type 1 diabetes

Long-standing diabetes damages the filters in the kidney, and the first sign is protein leaking into the urine. In one trial, children who already had that leak despite being on the standard kidney-protecting medicine were given carnosine on top. The amount of protein in their urine fell by more than half, a marker of damage to the kidney tubes fell, and their three-month blood sugar average came down — all of them further than in the children given a dummy capsule. Nobody has repeated this, and nobody followed the children long enough to see whether their kidneys did better in the end.

One randomised placebo-controlled trial: 90 children and young people with type 1 diabetes and kidney damage, already taking an ACE inhibitor, given 1 g a day or a matching dummy for 12 weeks, at one hospital in Cairo.

One small study

Source [12]

The brain, in the hours after a large dose

A brain scanner can read how much carnosine is in brain tissue. After a single large oral dose, the reading went up at one hour and was back to where it started by five hours. That is the only direct measurement in a living human brain, and it says the molecule can get there briefly. It says nothing about whether being there briefly changes anything.

Five people out of the sixteen in the dose-escalation study had magnetic resonance spectroscopy of the brain before and after dosing. Five people is small enough that the result should be read as a demonstration that the measurement is possible rather than as a finding about people in general.

One small study

Source [04]

Ageing cells in a dish

This is where the anti-ageing story comes from, and it is worth knowing exactly what it is. Human skin and lung cells grown in a laboratory dish divide a fixed number of times and then stop, looking flat and old. Adding carnosine to the liquid the cells live in kept them looking young for longer, and switching the carnosine off made them look old again. It did not abolish the hard ceiling on cell division — the cells still stopped in the end — although the number of divisions before they stopped was often larger. The concentration used was 20 to 50 millimoles per litre, which is enormously higher than anything a person can reach in their blood by swallowing capsules.

Two laboratory papers on cultured human fibroblast cell lines, from 1994 and 1999, by the same two authors. There is no living animal in either, and no person. No study has shown that carnosine lengthens life or slows ageing in a human being.

Only seen in a dish, not in a body

Source [25]Source [26]

Start with what did not work. Pooling five trials in 215 children with autism — the use carnosine first became known for — found no difference from a dummy capsule on either rating scale used: −2.57 points on the Gilliam scale (95% confidence interval −10.30 to 5.16, p = 0.52) and −0.88 on the Childhood Autism Rating Scale (−6.96 to 5.20, p = 0.78), and the reviewers concluded the data do not support its use. The largest psychiatric trial, 100 patients on 400 mg then 800 mg a day for 24 weeks, found no difference from a dummy on the negative-symptom score it was designed around. The most consistent positive signal is blood sugar: in 43 adults with prediabetes or type 2 diabetes on 2 g a day for 14 weeks, blood sugar 90 and 120 minutes after a sugary drink was 1.31 and 1.60 mmol/L lower than on a matching dummy (p = 0.02) with no change in insulin, and in 54 adults with type 2 diabetes on 1 g a day for 12 weeks fasting glucose was 13.1 mg/dL lower and HbA1c 0.6 percentage points lower than on placebo. One trial in 90 children with type 1 diabetes and kidney damage cut urinary protein from 91.7 to 38.5 mg/g creatinine against placebo. Two pooled analyses that produce friendlier headline numbers — HbA1c −0.36 standardised units across 8 trials in 377 people, and depression scores −0.79 in a review of 20 trials in 776 people, only 18 of which entered any pooled estimate — both mix carnosine trials together with beta-alanine trials without separating them, so neither figure can be read as a number about carnosine. Nothing has run longer than 24 weeks.

Two clocks run at different speeds. In the blood, everything happens in an afternoon: levels peak within about an hour of swallowing it, very little is left after four hours, and in the one brain-scanning study the brain reading was up at one hour and back to baseline by five. In the tissues, nothing happens quickly: raising the carnosine stored in muscle takes weeks, and the published trials of blood sugar, autism, schizophrenia and kidney damage all ran between 8 weeks and 24 weeks before measuring anything. Twenty-four weeks, in the 100-person schizophrenia trial, is the longest anyone has been given carnosine in a published randomised trial. There is no published data on what a year of it does, or two.

What was measuredWhat happenedIn whomHow it is known

Blood sugar after a standard sugary drink

1.31 mmol per litre lower at 90 minutes and 1.60 mmol per litre lower at 120 minutes on carnosine than on a matching dummy capsule (p = 0.02 for both), with the total sugar load over the two hours 3.30 mmol per litre lower (p = 0.04). Insulin did not differ between the groups. Nothing else the trial measured — body composition, calf muscle density, body measurements — differed either.

43 adults with prediabetes or type 2 diabetes, 30% women, 2 g a day for 14 weeks

Measured in people

Source [01]

Fasting blood sugar and the three-month blood sugar average in type 2 diabetes

Fasting blood sugar 13.1 mg per decilitre lower and the three-month average (HbA1c) 0.6 percentage points lower than on a dummy capsule; blood fats and a marker of sugar damage to proteins also came down further (all p < 0.05). Fasting insulin, insulin resistance, blood pressure, body mass index, total cholesterol and two of the three inflammation markers showed no difference between the groups.

54 adults with type 2 diabetes, 1 g a day as two 500 mg capsules, 12 weeks

Measured in people

Source [11]

Autism rating scales in children

No difference from a dummy capsule. On the Gilliam scale the difference between the groups was −2.57 points (95% confidence interval −10.30 to 5.16, p = 0.52), and on the Childhood Autism Rating Scale −0.88 points (−6.96 to 5.20, p = 0.78). A confidence interval that straddles zero this widely means the result is compatible with carnosine helping a little, doing nothing, or being slightly worse. The reviewers concluded that the data do not support using it.

215 children across 5 trials, 4 of them double-blind and placebo-controlled, 500 to 800 mg a day for 8 weeks to 2 months

Measured in people

Source [02]

The withdrawn, flattened side of schizophrenia

No difference between carnosine and a dummy capsule on the negative-symptom scale at 24 weeks, which was what the trial was built to test. One secondary measure, attention, separated at the higher dose (p = 0.023). A trial that misses the outcome it was designed around and reports one secondary measure instead is a trial that did not show what it set out to show.

100 patients with schizophrenia and prominent negative symptoms, 400 mg a day for 12 weeks then 800 mg a day to week 24

Measured in people

Source [10]

Blood sugar measures, pooled across trials — with a warning attached

Pooled across 8 trials in 377 people with prediabetes or type 2 diabetes, fasting blood sugar was lower than on a dummy by a standardised difference of −0.53 (95% confidence interval −0.75 to −0.31) and the three-month average by −0.36 (−0.59 to −0.12). The warning is that this pool mixes trials of carnosine with trials of beta-alanine and does not analyse the two separately, so the figure cannot be read as a number about carnosine. Insulin resistance did not shift significantly in the same pool.

377 adults with prediabetes or type 2 diabetes across 8 randomised trials of carnosine or beta-alanine

Studies in people disagree

Source [13]

Depression scores, pooled — with the same warning attached

Lower than on a dummy by a standardised difference of −0.79 (95% confidence interval −1.24 to −0.35) on a standard depression questionnaire, rated moderate-certainty by the reviewers. Quality of life rose by 0.65 points (0.00 to 1.30), which is a confidence interval that reaches exactly zero, and the reviewers rated that evidence low quality. As above, the pool mixes carnosine with anserine-carnosine blends and with beta-alanine, most of the trials ran under two months, and most carried a moderate to high risk of bias.

776 participants across 20 randomised trials of histidine-containing dipeptides, 18 of them pooled

Studies in people disagree

Source [14]

Executive-function tests in schizophrenia

Faster than on a dummy capsule on one part of a set-shifting test, and better on strategic efficiency with fewer repeated-mistake errors on a target-detection test. The other parts of the same set-shifting test, and every other cognitive test in the battery, showed no difference between the groups. Symptom scores stayed level in both arms.

75 adults with stable long-term schizophrenia, 2 g a day added to their usual treatment, 3 months

One small study

Source [15]

Protein leaking into the urine in children with diabetic kidney damage

Fell from 91.7 to 38.5 mg per gram of creatinine on carnosine, and the three-month blood sugar average fell from 8.2% to 7.4%, both significantly further than in the children on a dummy capsule (p < 0.001 against placebo). A marker of damage to the kidney tubes and a marker of oxidative damage both fell further too. No side effects attributed to carnosine were reported.

90 children and young people with type 1 diabetes and kidney damage already on an ACE inhibitor, 1 g a day, 12 weeks

One small study

Source [12]

Fatigue, pain and activity in Gulf War illness

Fatigue, pain, pain sensitivity, activity levels and quality-of-life scores — the things the trial set out to change — were no different on carnosine than on the dummy tablet. The working-memory task the trial's size was calculated around is not reported in this paper at all; the authors say those results will appear elsewhere. The one positive signal was a within-group one: the carnosine group's score on a pen-and-paper symbol test improved from its own starting point while the dummy group's did not, which the authors themselves say fell short of their conservative criteria. Diarrhoea associated with irritable bowel went from 20% of the carnosine group to 14%, while the dummy group stayed at about 19%.

34 veterans enrolled and 25 completing, 12 on carnosine and 13 on a dummy capsule, rising from 500 to 1,500 mg a day over 12 weeks

One small study

Source [27]Source [28]

The carnosine stored in muscle

Rose 65.8% from its own starting point after four weeks of swallowing L-carnosine, against 64.2% on the matched beta-alanine dose. The comparison here is between two supplements rather than against a dummy powder, and the two are not independent: the carnosine was cut apart in the blood first, and it was the beta-alanine released from it that the muscle used. Carnosine itself was not detectable in blood plasma in this study.

Small groups of healthy volunteers taking either beta-alanine or an amount of L-carnosine containing the same quantity of beta-alanine, for 4 weeks

One small study

Source [05]

At the amounts used in trials, 0.5 g to 2 g a day, little has been reported, but only one trial counted side effects against a control group and there carnosine came out at 30 per cent against 14 per cent on a dummy capsule over three months. A dose-escalation study in 16 healthy volunteers set the largest tolerable single dose at 10 g: nothing was reported at 4 g, one participant reacted at 6 g and three at 10 g, while at a single 15 g dose 77 per cent had a problem — most often headache (43.5 per cent), feeling sick (21.7 per cent) and pins and needles in the skin (21.7 per cent), the last being the tingling beta-alanine is known for, which appears here because carnosine breaks down into it. One transient rise in liver blood tests was recorded in the Gulf War illness trial, where 7 of the 9 dropouts were in the carnosine arm. The US regulator's public complaints database holds only 7 reports naming a carnosine product, far too few to be a safety record in either direction, and four of those seven people were taking eleven to fifteen other supplements at the same time. No published trial has enrolled anyone pregnant or breastfeeding, and because two randomised trials lowered blood sugar against placebo, anyone on diabetes medicine should tell a doctor before adding it.

EffectWhat it feels likeHow often

Headache

This comes from a dose-escalation study in healthy volunteers with no dummy-capsule group, so there is nothing to compare the rate against — some proportion of any group of people reports a headache in any given week. What the study does establish is a ceiling: the authors set the largest tolerable single dose at 10 g, because at 15 g the rate of problems became, in their words, unacceptably high.

The commonest problem at high single doses: reported by 43.5% of the participants at a single 15 g dose. At 4 g nobody reported anything at all, at 6 g one person out of 16 did, and at 10 g three out of 15 did.

Source [04]

Feeling sick

Again from an escalation study with no dummy comparison. The amounts that produced it are far above the 1 to 2 g a day used in the trials of blood sugar, autism and schizophrenia, and no adverse events at all were recorded in the four people who took 5 g twice a day for four weeks.

21.7% of participants at a single 15 g dose, and 1 of 16 at 6 g. Not separately reported in the long-term trials.

Source [04]

Pins and needles in the skin

This is the tingling, prickling sensation on the face, neck and hands that beta-alanine is famous for among people who take it before training. It turns up after a large carnosine dose for the obvious reason: carnosine is cut in half in the blood, and one of the halves is beta-alanine. It is harmless and it passes. Sports nutrition guidance describes it as the only reported side effect of beta-alanine, appearing above about 800 mg taken in one go in a fast-release form.

21.7% of participants at a single 15 g dose.

Source [04]Source [20]

Any side effect at all, counted against a dummy capsule

This is the only figure on this page that compares side effects between a carnosine group and a control group. It is one trial of 75 people, and the difference is the kind that a trial this size cannot separate cleanly from chance. It is here because it is the only number of its kind, not because it settles anything.

30% of the carnosine group against 14% of the dummy group over three months, in the one trial that reported it that way. Blood test results stayed within acceptable ranges in both groups.

Source [15]

A temporary rise in liver blood tests

Worth noting alongside it: 7 of the 9 people who dropped out of that 34-person trial were in the carnosine arm, and the reasons given included not feeling any better. A dropout is not a side effect, but a lopsided dropout pattern is the kind of thing a larger trial would need to explain.

One event in the Gulf War illness trial, graded as the mildest category and resolving on its own.

Source [27]

These are the amounts the studies below gave their participants, and they are here as facts about those studies. They are not a recommendation, not a starting point and not a range to pick from. What a person should take, if anything, is a question for a doctor who knows them.

Ali et al. 2025, a dose-escalation study in healthy volunteers to find the ceiling

Single doses of 4 g, 6 g, 10 g and 15 g, swallowed, with two weeks between each — a further 20 g step was in the protocol but was never given, because the study was stopped at 15 g; separately, 5 g twice a day

Single doses two weeks apart; the twice-daily arm ran 4 weeks in 4 people

Source [04]

Hariharan et al. 2024, blood sugar control in adults with prediabetes or type 2 diabetes

2 g a day, swallowed, against a matching dummy capsule

14 weeks

Source [01]

Houjeghani et al. 2018, blood sugar and blood fats in type 2 diabetes

1 g a day as two 500 mg capsules, against a matching dummy

12 weeks

Source [11]

Elbarbary et al. 2018, kidney damage in children with type 1 diabetes

1 g a day on top of an ACE inhibitor, against a matching dummy

12 weeks

Source [12]

Chengappa et al. 2012, thinking tests in schizophrenia, added to usual treatment

2 g a day, against a matching dummy

3 months

Source [15]

Tharoor et al. 2023, negative symptoms in schizophrenia — the longest published dosing period

400 mg a day for the first 12 weeks, then 800 mg a day, against a matching dummy

24 weeks

Source [10]

Chez et al. 2002, the first autism trial

800 mg a day, against a matching dummy

8 weeks

Source [23]

Baraniuk et al. 2013, Gulf War illness, with the dose stepped up during the trial

500 mg a day for 4 weeks, then 1,000 mg a day for 4 weeks, then 1,500 mg a day, against a cellulose dummy tablet

12 weeks

Source [27]

Harris et al. 2006, the muscle-loading study — the one place a carnosine dose was set by chemistry rather than by convention

An amount of L-carnosine containing the same quantity of beta-alanine as the study's 6.4 g a day beta-alanine arm, swallowed in divided doses

4 weeks

Source [05]

What people report

• Whether swallowing carnosine does anything that swallowing the cheaper beta-alanine it breaks down into would not do — no trial has ever compared the two head to head for any outcome except the amount stored in muscle.
• Whether the blood sugar effect lasts, or matters: the longest trial ran 14 weeks, measured a sugar-drink test, and did not follow anyone to see whether they developed diabetes or avoided it.
• Whether the people whose blood breaks carnosine down fastest get any effect at all — the enzyme varies severalfold between individuals, and no trial has sorted its participants by it.
• What a year of it does, or two: 24 weeks is the longest anyone has been given carnosine in a published randomised trial.
• Whether the kidney result in 90 children with type 1 diabetes repeats — it rests on one trial at one hospital, and nobody has reproduced it.
• Whether it is safe in pregnancy or breastfeeding; no published trial has enrolled anyone pregnant or breastfeeding.
• Whether it interacts with any medicine, including the diabetes medicines several trial participants were already taking.
• Whether N-acetylcarnosine eye drops do anything for cataract: the only two trials came from the patent holder's own company, and a Cochrane review could not obtain enough information to assess either one.
• Whether the depression and blood sugar figures that get quoted actually belong to carnosine: both pooled analyses mix carnosine trials with beta-alanine trials and do not separate them.
• What a given capsule contains — this is regulated as a food, no independent analysis of retail carnosine products was found for this entry, and every trial used its own supplied material.
• Whether it does anything to the ageing process in a living human being; the entire ageing claim rests on two 1990s papers on cells in a dish.
• What became of a 47-person trial in bipolar disorder that finished in December 2007 and has posted no results, and of an 84-person cardiometabolic trial in Melbourne whose registry entry has not been updated since its 2020 completion date.
• Whether it helps in peripheral artery disease, where a 144-person phase 1/2 trial started recruiting in May 2025 and is not due to finish until October 2028.

Does carnosine actually work?

For one thing, in one group of people, the controlled evidence points in a consistent direction: blood sugar after a sugary drink came down further on carnosine than on a dummy capsule in adults with prediabetes or early type 2 diabetes, in two separate randomised trials. For the things it is mostly bought for — slowing ageing, improving memory, helping children with autism — it has either failed or has never been tested properly. The pooled result across five autism trials in 215 children shows no difference from a dummy, and the largest schizophrenia trial missed the outcome it was built around.

Source [01]Source [02]Source [10]

Carnosine vs beta-alanine — which should I take?

They are not two versions of the same thing, but they are closely related: carnosine is beta-alanine joined to histidine, and swallowing carnosine feeds you beta-alanine after your blood takes it apart. For raising the carnosine stored in muscle for hard, short efforts, beta-alanine is the half that has actually been studied for it, taken daily over a period of weeks, and sports guidance states outright that oral carnosine is an inefficient way to do it because it is broken down before it gets there. If the goal is anything else on this page, beta-alanine has not been tested for it.

Source [05]Source [18]Source [20]

Do carnosine eye drops work for cataracts?

There is no convincing evidence that they do. The drops sold for this contain N-acetylcarnosine, a modified form of carnosine, and a Cochrane review looked for trials of them in 2017. It found two, enrolling 114 people aged 55 to 80 between them, and could not get enough information about how either was designed or run to include them at all — the reviewers wrote to the author and got no reply. Both trials came from the same researcher, and the company that funded them holds the worldwide patent on using N-acetylcarnosine for eye disorders and markets the product. The review's conclusion is that there is currently no convincing evidence that the drops reverse a cataract or stop one getting worse.

Source [06]Source [29]

Is zinc carnosine the same as carnosine?

No. Zinc carnosine, generic name polaprezinc, is carnosine bound to zinc, and it is a different substance with a different molecular formula and its own evidence. It was developed as a stomach-ulcer drug and has been an approved prescription medicine in Japan since 1994. It is sold in Europe and the US as a supplement for the stomach lining. Nothing on this page about carnosine's effects on blood sugar, thinking or kidneys is evidence about zinc carnosine, and the reverse is equally true.

Source [07]Source [09]

Is carnosine the same as carnitine?

No, and the two get confused constantly because the names are one letter apart. Carnosine is two amino acid building blocks joined together, formula C9H14N4O3, and it works as a buffer and antioxidant in muscle and brain. Carnitine is a completely different molecule, formula C7H15NO3, involved in moving fat into the part of the cell that burns it. Nothing on this page applies to carnitine.

Source [09]Source [30]

What are the side effects of carnosine?

At the amounts used in trials — 0.5 g to 2 g a day — very little has been reported, though only one trial counted side effects against a dummy group, and there carnosine came out at 30% against 14%. At much larger single doses the picture is clearer: in a study that pushed the dose up step by step, nobody reported anything at 4 g, but at a single 15 g dose 77% of participants had a problem, most often headache, feeling sick, and pins and needles in the skin. Ten grams in one go was set as the largest tolerable single dose.

Source [04]Source [15]

Is carnosine legal, and is it banned in sport?

It is sold openly as a food supplement rather than as a medicine, and it needs no prescription. The US regulator's own database of approved medicines returns no product with carnosine as an active ingredient, and this register found no European marketing authorisation for it either. It is not banned in sport: neither carnosine nor beta-alanine appears anywhere in the World Anti-Doping Agency's 2026 Prohibited List, which was searched word by word for this entry on 4 August 2026.

Source [08]Source [16]

Can you get enough carnosine from food?

Carnosine is one of the reasons meat is meat — it was first isolated from it in 1900 and it is concentrated in muscle. A 200 g portion of cooked minced beef in one study contained about 250 mg of carnosine, and eating it put a measurable amount into the blood within 15 minutes. That is roughly a quarter of the daily amount used in several of the trials on this page, from one meal. Vegetarians measured 26% less carnosine in the calf muscle than meat-eaters, so diet does move the stores, but only somewhat.

Source [17]Source [19]Source [31]

Does carnosine slow ageing?

Nobody has tested that in a person. The anti-ageing claim comes from two laboratory papers from the 1990s in which human cells grown in a dish kept a youthful appearance for longer when carnosine was added to the liquid they lived in — at concentrations far above anything a body can reach — and reverted to looking old when it was taken away. Even in the dish the cells still hit their ceiling and stopped dividing, although they often managed more divisions before they did. No trial has measured lifespan, health span or any ageing outcome in a human being.

Source [25]Source [26]

Where can you buy carnosine, and does it need a prescription?

It is sold openly as a food supplement in capsules, usually 500 mg, in pharmacies, health shops and online, and it needs no prescription anywhere because no regulator has approved it as a medicine. That also means nobody checks what is in a given tub before it is sold, and no trial on this page tells you anything about a specific brand: each one bought its own material and tested that. The products named in the US regulator's complaints database are ordinary high-street supplement brands.

Source [16]Source [32]

1. [01]Hariharan R et al. Carnosine supplementation improves glucose control in adults with pre-diabetes and type 2 diabetes: a randomised controlled trial, 43 adults, 2 g/day, 14 weeks (Nutr Metab Cardiovasc Dis 2024;34(2):485-96) (2024)
2. [02]Abraham DA et al. Effect of L-carnosine in children with autism spectrum disorders: systematic review and meta-analysis of randomised controlled trials, 5 trials, 215 participants, no difference from placebo (Amino Acids 2021) (2021)
3. [03]Everaert I et al. Low plasma carnosinase activity promotes carnosinemia after carnosine ingestion in humans, 25 healthy adults, 8 responders (Am J Physiol Renal Physiol 2012;302(12):F1537-44) (2012)
4. [04]Ali AN et al. Dietary carnosine supplementation in healthy human volunteers: a safety, tolerability, plasma and brain concentration study, 16 volunteers, maximum tolerated single dose 10 g (Nutrients 2025;17(13):2130) (2025)
5. [05]Harris RC et al. The absorption of orally supplied beta-alanine and its effect on muscle carnosine synthesis in human vastus lateralis (Amino Acids 2006;30:279-89) (2006)
6. [06]Dubois VDJP, Bastawrous A. N-acetylcarnosine (NAC) drops for age-related cataract: no convincing evidence; both candidate trials from the patent holder's company and unassessable (Cochrane Database Syst Rev 2017, CD009493) (2017)
7. [07]Li M et al. Recent advances on polaprezinc for medical use (review): zinc complex of L-carnosine, first approved in Japan in 1994 (Exp Ther Med 2021;22(6):1445) (2021)
8. [08]WADA: World Anti-Doping Code International Standard – Prohibited List 2026 (in force 1 January 2026); neither carnosine nor beta-alanine appears anywhere in the document (2026)
9. [09]PubChem: carnosine, C9H14N4O3, 226.23 g/mol, CID 439224, CAS 305-84-0, containing beta-alanine which has no one-letter code
10. [10]Tharoor H et al. L-carnosine add-on in schizophrenia — 100 patients, 24 weeks (2023)
11. [11]Houjeghani S et al. L-carnosine supplementation in type 2 diabetes — 54 adults; fasting glucose down 13.1 mg/dL and HbA1c down 0.6 percentage points (2018)
12. [12]Elbarbary NS et al. The effect of 12 weeks carnosine supplementation on diabetic nephropathy in children — 90 children; albumin-creatinine ratio 91.7 to 38.5 mg/g (2018)
13. [13]Carnosine supplementation and glycaemic control: a systematic review and meta-analysis, BMC Endocr Disord 2025 — HbA1c −0.36 across 8 trials in 377 people (2025)
14. [14]Kabthymer RH et al. Carnosine supplementation and depressive symptoms: a systematic review and meta-analysis, 2024 — 20 randomised trials in 776 people, 18 entering the pooled estimate (2024)
15. [15]Chengappa KNR et al. A randomized, placebo-controlled, add-on trial of L-carnosine in schizophrenia, 2012 — adverse events 30% against 14% (2012)
16. [16]openFDA — carnosine returns NOT_FOUND as an approved US drug ingredient, and the food adverse-event database holds 7 reports naming a carnosine product
17. [17]Park YJ, Volpe SL, Decker EA. Quantitation of carnosine in human plasma after dietary consumption of beef, 18 adults (J Agric Food Chem 2005;53(12):4736-9) (2005)
18. [18]Cesak O et al. Carnosine and beta-alanine supplementation in human medicine: narrative review and critical assessment (Nutrients 2023;15(7):1770) (2023)
19. [19]Everaert I et al. Vegetarianism, female gender and increasing age, but not CNDP1 genotype, are associated with reduced muscle carnosine levels in humans, 149 subjects (Amino Acids 2011;40(4):1221-9) (2011)
20. [20]Trexler ET et al. International Society of Sports Nutrition position stand: beta-alanine (J Int Soc Sports Nutr 2015;12:30) — states that oral carnosine is an inefficient way to raise muscle carnosine (2015)
21. [21]de Courten B et al. Effects of carnosine supplementation on glucose metabolism: pilot clinical trial, 30 adults with overweight or obesity (Obesity 2016;24(5):1027-34) (2016)
22. [22]Hsiao et al. The effects of carnosine on cognitive function and mental health: systematic review and meta-analysis of 13 studies (Nutrients 2026;18(9):1385; supported by Natural Alternatives International, Inc.) (2026)
23. [23]Chez MG et al. Double-blind, placebo-controlled study of L-carnosine supplementation in children with autistic spectrum disorders, 31 children, 8 weeks (J Child Neurol 2002;17(11):833-7) (2002)
24. [24]Mehrazad-Saber M et al. Effects of l-carnosine supplementation on sleep disorders and disease severity in autistic children: randomized controlled clinical trial, 43 children, 500 mg a day, 2 months (Basic Clin Pharmacol Toxicol 2018;123:72-77) (2018)
25. [25]McFarland GA, Holliday R. Retardation of the senescence of cultured human diploid fibroblasts by carnosine (Exp Cell Res 1994;212(2):167-75) (1994)
26. [26]McFarland GA, Holliday R. Further evidence for the rejuvenating effects of the dipeptide L-carnosine on cultured human diploid fibroblasts (Exp Gerontol 1999;34(1):35-45) (1999)
27. [27]Baraniuk JN et al. Carnosine treatment for Gulf War illness: a randomized controlled trial (Glob J Health Sci 2013;5(3):69-81) (2013)
28. [28]NCT00810368, carnosine versus placebo in Gulf War illness, 33 participants, completed July 2012 (2012)
29. [29]The same review's study tables (full text), recording Innovative Vision Products, Inc. as funder and holder of the worldwide patent for the ophthalmic use of N-acetylcarnosine, and NAC as marketed by the lead author of both trials (2017)
30. [30]PubChem: L-carnitine, C7H15NO3, 161.20 g/mol, CID 10917
31. [31]Boldyrev AA, Aldini G, Derave W. Physiology and pathophysiology of carnosine (Physiol Rev 2013;93(4):1803-45) (2013)
32. [32]openFDA food, dietary supplement and cosmetic adverse event reports (CAERS), products named carnosine, retrieved 4 August 2026: 7 reports, all classified as dietary supplements