L-carnitine is a compound your body makes and also gets from food, especially red meat, and it helps move fatty acids into your cells’ energy-producing mitochondria. Human studies show small to moderate effects on some outcomes, including blood sugar, sperm motility and weight, and mixed results for heart outcomes, exercise and memory. The NIH says healthy people do not need to get carnitine from food or supplements, and some studies raise questions about heart-related safety.
Quick facts
| Item | What the research shows |
|---|---|
| Forms | L-carnitine and acetyl-L-carnitine are the two main supplement forms. Propionyl-L-carnitine and L-carnitine tartrate appear in some studies. |
| Doses used in studies | From 0.25 g/day up to 6 g/day, depending on the trial and condition (details below). Supplements on the market range from about 3 mg to 5,000 mg. |
| Typical study length | A single dose or one week up to 3 years. Many trials run 8 weeks to 12 months. |
| Main studied uses | Blood sugar and insulin resistance, heart disease, male fertility, kidney dialysis, weight loss, exercise performance, dementia, leg pain from poor circulation, knee osteoarthritis, polycystic ovary syndrome (PCOS) |
| Absorption | About 14% to 18% of supplemental L-carnitine is absorbed, versus about 63% to 75% of carnitine from food, according to the NIH Office of Dietary Supplements fact sheet. |
What does your body already do with carnitine?
Your liver, kidneys and brain make carnitine from the amino acids lysine and methionine. About 95% of the body’s carnitine sits in heart and skeletal muscle. The NIH fact sheet says endogenous production meets the needs of healthy people, and that is why no official daily requirement exists.
A typical omnivorous diet supplies about 24 to 145 mg a day, while a vegan diet supplies about 1.2 mg. A 3-ounce serving of cooked beef steak has roughly 42 to 122 mg, and chicken breast has about 2 to 4 mg. Deficiency happens in genetic disorders, in premature infants, in kidney failure and dialysis, and with some medicines.
What does the human evidence show?
The studies below are ordered roughly from the largest bodies of trials to the smallest. For how to weigh them, see our guides on meta-analyses and systematic reviews and evaluating evidence quality. Many results below are changes in lab markers rather than in how people feel or live, which our guide to surrogate outcomes explains.
Blood sugar and insulin resistance: modest improvements in lab markers
A 2023 meta-analysis of 41 randomized trials (2,900 adults, most with type 2 diabetes, obesity, PCOS or fatty liver) found that 0.25 to 4 g/day for 2 to 52 weeks lowered fasting blood glucose, insulin resistance and HbA1c, but not serum insulin. It is summarized in the NIH fact sheet. A narrower meta-analysis of four trials in 284 adults with type 2 diabetes (2 or 3 g/day for 12 to 52 weeks) found lower fasting glucose and cholesterol but no change in HbA1c or triglycerides.
A 2025 meta-analysis in women with overweight or obesity pooled eight studies (nine trial arms). Carnitine lowered fasting blood sugar by 2.93 mg/dL, fasting insulin by 3.54 μU/mL and HOMA-IR (an insulin resistance score) by 0.69 versus placebo. Effects were stronger at 2,000 mg/day or more and at 12 weeks or longer. The authors called the improvements modest. The NIH says larger trials are needed to show whether carnitine lowers the risk of diabetes or its complications.
Heart attack, coronary disease and heart failure: some benefit signals, mixed overall
The NIH fact sheet describes a meta-analysis of 13 trials in 3,629 adults after a heart attack, using 2.7 g/day for 5 days up to 6 g/day for 12 months. L-carnitine was linked to lower all-cause death, fewer dangerous heart rhythms and less new angina. It did not change heart failure or repeat heart attack rates. A 2026 systematic review of 21 studies reached the same pattern and called results on heart failure and reinfarction inconsistent. Its authors say large, long-term randomized trials are still needed.
A meta-analysis of 17 trials in 1,625 adults with chronic heart failure (1 to 6 g/day, 7 days to 3 years) found improved ejection fraction (4.14%), stroke volume (8.21 mL) and cardiac output (0.88 L/min) compared with routine treatment alone. Death rates and timed walking test results did not change. These trials added carnitine to usual care; none showed it replacing it.
Male fertility and testosterone: better sperm markers, unclear for pregnancy
Does L-carnitine raise testosterone? One 2025 meta-analysis of seven randomized trials in men with idiopathic (unexplained) infertility found L-carnitine raised serum testosterone, with a p-value under 0.001. It also raised sperm concentration, normal shape and motility. Only four of the seven trials tested L-carnitine, the abstract does not give the size of the testosterone change, and the men all had fertility problems. It does not show that carnitine raises testosterone in healthy men.
Other reviews fit the sperm findings:
- A review of three trials in 201 men (1 to 3 g/day for 2 to 6 months) found 7.84% better motility and 4.91% better morphology, but no change in sperm concentration (NIH summary).
- A 2022 Cochrane review of six trials in 1,089 men found higher motility and concentration at some time points, but no difference in live birth or pregnancy rates.
- A 2025 network meta-analysis of 16 studies ranked L-carnitine best for sperm motility and coenzyme Q10 best for sperm concentration. Its authors call for larger, higher-quality studies.
Kidney dialysis: a real use under medical care, with limited trials
Many people on hemodialysis become carnitine-insufficient. A meta-analysis cited by the NIH concluded that carnitine might help manage anemia but not blood fats, and that effects on exercise capacity and heart disorders were inconclusive. Most of the studies were small and not double-blind. In children on dialysis, a 2026 systematic review of nine studies (194 patients, three of them randomized) found partial benefits: higher hemoglobin, a lower required dose of anemia medicine, lower ApoB and better cardiac function in some studies. The authors rated study quality as not high and called for better trials.
Weight loss and belly fat: a small average effect, nothing specific to belly fat
Does L-carnitine get rid of belly fat? The research found does not measure belly fat. It measures body weight. A 2016 meta-analysis of nine randomized trials (911 people; mostly 1.8 to 4 g/day for 30 to 360 days) found people on carnitine lost an average of 1.33 kg more than those on placebo, regardless of dose or duration. The NIH notes weight was a secondary outcome in most studies and results were equivocal.
One trial in 258 adults with uncontrolled type 2 diabetes added 2 g/day of L-carnitine to the prescription drug orlistat (360 mg/day) for a year. They lost 11.3 kg versus 9.5 kg on orlistat alone. That is one trial in people with diabetes, and it tested carnitine as an add-on, not as a substitute. In another trial, 2 g/day alone for 6 months did not change weight in 94 overweight people with newly diagnosed type 2 diabetes.
Exercise and athletic performance: mixed
The NIH calls the evidence mixed. In a 24-week trial of 14 recreational athletes using L-carnitine tartrate (2.0 g twice a day) with carbohydrate, perceived exertion was lower on a bike test. In a 12-week trial of 24 men (1 g twice daily), muscle carnitine rose about 13% in vegetarians but not omnivores, and there was no change in VO2max or performance.
A review of 11 trials in 251 trained athletes and another covering 17 studies in 237 active adults found benefits on lactate, recovery or VO2max in some studies and none in others. A systematic review of 11 trials in 203 untrained or active adults found no study showing improved moderate-intensity performance.
PCOS fertility: positive trials, small number
In a randomized trial of 170 women under 35 with PCOS taking the fertility drug clomiphene citrate, adding 3 g/day of L-carnitine was linked to 64.4% higher ovulation rates, thicker uterine lining and more pregnancies. In a 3-month trial of 280 women with PCOS taking 3 g, cycle regularity, ovulation and pregnancy rates improved, but miscarriage rates did not differ. The NIH says more research is needed.
Memory and dementia: mixed and not supported for routine use
These studies used acetyl-L-carnitine. A 2003 meta-analysis of 21 trials (1,204 adults with mild cognitive impairment or mild Alzheimer’s, 1.5 to 3 g/day for 3 to 12 months) favored the supplement. A 2003 Cochrane review of 15 trials found symptom benefits at 12 and 24 weeks but not at 52 weeks, and no effect on dementia severity or daily function. The Cochrane authors concluded routine clinical use was not justified.
Leg pain from poor circulation and knee osteoarthritis: inconsistent
A review of three trials of propionyl-L-carnitine (2 g/day for 4 to 6 months; 234 people versus 222 on placebo) found better walking time in one trial and no benefit in the other two. In knee osteoarthritis, an 8-week trial in 69 women (250 mg three times daily) found lower inflammation markers and pain scores, but a 12-week trial in 76 women with obesity (1 g/day) found no change in pain, stiffness or function. For premature newborns on intravenous nutrition, a Cochrane review of six trials did not support carnitine for fat use or weight gain.
What doses did researchers use?
These are amounts used in studies. They are not recommendations.
| Use studied | Form | Amounts and length in studies |
|---|---|---|
| Blood sugar (41-trial review) | L-carnitine | 0.25 to 4 g/day, 2 to 52 weeks |
| After heart attack | L-carnitine | 2.7 g/day for 5 days up to 6 g/day for 12 months |
| Chronic heart failure | L-carnitine | 1 to 6 g/day, 7 days to 3 years |
| Male infertility | L-carnitine or acetyl-L-carnitine | 1 to 3 g/day, 2 to 6 months |
| Weight loss | L-carnitine | 1.8 to 4 g/day for 30 to 360 days (one trial 15 mg/kg/day for 182 days) |
| Exercise | L-carnitine, L-carnitine tartrate | 1 to 4 g, single dose to 6 months |
| Dementia | Acetyl-L-carnitine | 1 to 3 g/day, 12 to 52 weeks |
| Leg pain (circulation) | Propionyl-L-carnitine | 2 g/day, 4 to 6 months |
For diagnosed primary or secondary carnitine deficiency, the NIH says high doses of 20 to 200 mg/kg/day have been used, which is a medical treatment decided by a clinician.
Does L-carnitine work immediately?
Most results above came from weeks to months of daily use. Sperm effects showed at some time points such as 3 months, the 2025 blood sugar review found stronger effects at 12 weeks or longer, and one review of insulin resistance found 12-month benefits exceeded 3-month ones. Single-dose exercise studies exist, but their results were mixed. No study here shows a reliable immediate effect.
Is it safe? Side effects, heart questions and medicine interactions
The NIH fact sheet says carnitine has no established tolerable upper intake level. Doses of about 3 g/day of supplements can cause nausea, vomiting, abdominal cramps, diarrhea and a fishy body odor. It can cause muscle weakness in people with uremia and seizures in people with seizure disorders.
| Issue | What the NIH reports |
|---|---|
| Stomach and body odor | Nausea, vomiting, cramps, diarrhea, fishy odor at about 3 g/day |
| Uremia (kidney failure) | Can cause muscle weakness |
| Seizure disorders | Can cause seizures |
| Gut bacteria and TMAO | Intestinal bacteria turn unabsorbed carnitine into TMAO, a substance linked to cardiovascular risk. The effect appears stronger in meat eaters than in vegans or vegetarians. The NIH says the meaning of this is not well understood. |
| Pivalate-conjugated antibiotics (such as pivampicillin) | Long-term use can deplete carnitine. No illness from this has been described in the sources the NIH cites. |
| Valproic acid, phenobarbital, phenytoin, carbamazepine | These anticonvulsants lower blood carnitine. Valproic acid can also affect the liver and raise ammonia. |
Two heart-related studies explain the NIH’s caution. A 2013 study of 2,595 adults found L-carnitine is turned into TMAO by gut bacteria, and higher blood carnitine went with higher risk of heart and artery disease only in people who also had high TMAO. A 2022 trial in 157 people aged 58 to 75 with metabolic syndrome (1 g twice a day for 6 months) found no difference in total plaque volume, but total and LDL cholesterol were higher on L-carnitine. Carotid artery narrowing was 9.3% greater in men who ate less red meat and started with less narrowing.
Is L-carnitine safe for your kidneys?
The NIH does not report kidney damage from carnitine. It does say kidney disease changes how the body handles carnitine, and that carnitine can cause muscle weakness in uremia. Dialysis patients are sometimes treated for deficiency, but under medical supervision, often by injection. If you have kidney disease, ask your kidney doctor before taking it.
What happens if you take it every day?
Trials gave it daily for weeks to months, and a few for up to 3 years. The NIH reports the digestive effects above at around 3 g/day. Daily intake raised muscle carnitine in vegetarians in one 12-week study. How daily use over many years affects the heart is not settled, given the TMAO findings.
Who should talk to a clinician first?
The NIH says people who take any medicine regularly should discuss carnitine with their healthcare provider. Based on the safety findings, raise it first if you:
- have kidney disease or are on dialysis;
- have a seizure disorder or take valproic acid, phenobarbital, phenytoin or carbamazepine;
- take pivalate-conjugated antibiotics;
- have heart disease, metabolic syndrome or high cholesterol;
- take orlistat or clomiphene, which were combined with carnitine in some trials;
- are considering it for a child with a medical condition.
Carnitine should not replace any prescribed medicine, and any change to a medicine is a decision for your prescriber.
Is it good for your health, and is it better than creatine?
For healthy people, the NIH says the body makes enough. Benefits in studies were mostly in people with specific conditions, and many were changes in lab values. No study here compares L-carnitine with creatine, so a ranking cannot be made from this research. If you are deciding between them for a goal, ask a clinician or sports dietitian.
What is still unknown?
- Whether carnitine lowers diabetes risk or its complications. Larger trials are needed.
- Whether it prevents heart events in people without heart disease, and what the TMAO findings mean in practice.
- Whether it helps memory or dementia. Results were more favorable in older trials than recent ones.
- Whether sperm improvements lead to more pregnancies. Live birth and pregnancy rates did not change in the Cochrane review.
- Whether it changes testosterone in men without fertility problems.
- How well acetyl-L-carnitine is absorbed. The NIH says this is not well studied.
- Effects on belly fat, which studies did not measure.
How do you read a label that lists L-carnitine?
- Check the form. L-carnitine, acetyl-L-carnitine, propionyl-L-carnitine and L-carnitine tartrate are different compounds, and the trials above used specific ones. Results for one form do not automatically apply to another.
- Check the amount per serving. Label amounts range from about 3 mg to 5,000 mg. Compare the number with the amounts in the table above and remember that about 14% to 18% of supplemental L-carnitine is absorbed.
- Check how the amount is listed. If the label gives only a combined total for a “blend”, you cannot tell how much carnitine you are getting. Ask the maker for the exact amount of each ingredient.
- Check the serving size and number of servings. Side effects were reported at about 3 g/day, so add up everything you take.
- Watch for claims. Terms like “fat burner” or “detox” go beyond what the trials measured. Our guide to ingredient evidence versus marketing claims shows how to compare a claim with the research.
By ClinicalStudyConnect.com Research Desk