Home
Ingredient Beauty Benefits

Other · Carnitine

L-carnitine: sebum went down — but it was the losing arm in its own trial

In a four-week trial splitting 90 people into three arms, 2% l-carnitine cream significantly reduced sebum (P < .01). But it took the longest to control oil, and the arm combining it with EGCG beat it significantly (P = .009). And there was no arm that got the base alone.

Last checked Sep 19, 2026

In pregnancy
Not established
Irritation
Minimal
Daytime use
✓ Fine

Not a measurement — reported irritation grouped into five steps. It varies between people.

Korean classification Being checkedThe regulatory class has not yet been confirmed against a primary document. It will be filled in once it is.

Carnitine is the carrier molecule the body uses to burn fat — it ferries fatty acids into mitochondria. You may know it better as an oral supplement; this page is about applying it to skin.

What it is expected to do topically is reduce sebum production, by touching fat metabolism inside the sebaceous gland.

This page has only one trial, and that trial also appears in full on the EGCG page.

There were three arms in one trial, and this ingredient was the worst-performing of them. So this page and the EGCG page read the same data from different angles.

We wrote the same limitation onto both pages. Dropping a limitation only on the page where it is inconvenient is itself a bias.

Contents
  1. Sebum fell significantly — in all three arms
  2. Within the same trial it was the slowest, and it lost to the combination
  3. We found no other human trial of topical carnitine

What is established

The same ingredient at a different strength, in a different formula, is a different product. These are the conditions the studies used.

Moderate evidenceConcentration studied 2% · cream

Sebum fell significantly — in all three arms

A 2023 randomised clinical trial in Thailand, in people with seborrhea — faces that turn greasy quickly.

Ninety people were split into three arms and applied cream for four weeks. All three creams put an active into the same dimethicone and glycerin moisturising base.

  • 2% l-carnitine
  • 5% EGCG
  • 2% l-carnitine + 5% EGCG

Sebum level, stratum corneum capacitance and transepidermal water loss (TEWL) were measured at weeks 0, 1, 2 and 4.

Result: sebum fell significantly from baseline in all three arms (P < .01). The other objective parameters and subjective outcomes improved in all three as well.

The instrument endpoint is what is valuable. Sebum comes out as a number on a sebumeter — a reading, not a feeling. It is why oiliness is one of the few concerns on this site checked with numbers.

But there is a large hole here.

All three arms received a cream with an active in it. There was no arm applying the base alone.

The base is dimethicone and glycerin. Both do something to skin on their own, and the act of applying a moisturising cream is itself a matter of wiping off surface oil and covering it again.

So 'sebum fell significantly' cannot distinguish between:

  1. Carnitine reduced sebum production
  2. The base alone would have reduced it as much
  3. It fell because season and humidity changed over four weeks

The 1983 trial on the lactic acid page handled exactly this — a plain emollient was included as a comparator and all three improved, and the ingredient's contribution only emerged after stopping.

The same paragraph appears on the EGCG page.

  • Randomised controlled trial in people · 90 participants · 4 weeks · Compared with another active ingredient · Funding not declared · Endpoint instrument measurement PMID 37194909
Moderate evidenceConcentration studied 2% · cream

Within the same trial it was the slowest, and it lost to the combination

The between-arm comparisons from that trial. They go against this ingredient, which is all the more reason to write them plainly.

One — it was the slowest.

The median time to oil control was longer in the l-carnitine group.

All three arms eventually reduced sebum, but this one took longest. In a four-week trial that is a practical difference — for someone using it, 'when does it start working' matters as much as 'does it work eventually'.

Two — it lost significantly to the combination.

The combination group had significantly greater anti-sebum efficacy than the l-carnitine group (P = .009).

What kind of comparison this is, precisely.

The combination arm is l-carnitine 2% plus EGCG 5%. That is the l-carnitine arm with EGCG added. So this is an add-on comparison, and the difference belongs to EGCG. The kojic acid page has a design of the same shape.

Read the other way round, it also means adding something to l-carnitine makes it better. This ingredient alone is not the ceiling.

And one comparison is missing from the abstract. There is no figure comparing the l-carnitine arm against the EGCG arm head to head. The only significant pair reported is combination versus l-carnitine.

So this abstract alone cannot support 'EGCG is better than l-carnitine'. What is established goes as far as the two together beat l-carnitine alone.

  • Randomised controlled trial in people · 90 participants · 4 weeks · Compared with another active ingredient · Funding not declared · Endpoint instrument measurement PMID 37194909
Not establishedConcentration studied — Not established

We found no other human trial of topical carnitine

The only human trial of topical l-carnitine we found on PubMed is the one above.

So everything on this page rests on 90 people, four weeks, one trial. And within that trial this ingredient was the worst-performing of three arms.

Do not confuse it with oral carnitine. There is far more data on carnitine as a supplement, but taking it and applying it are different matters. The collagen page on this site draws the same distinction — a different route needs its own evidence.

And the 'it burns fat' logic needs care when carried to skin. That carnitine is used in fatty acid transport is a fact of biochemistry, but that this happens in the sebaceous gland when you apply it to skin has to be established separately. The exosome and PLLA pages make the same distinction repeatedly.

Separating what is and is not established:

  • Sebum fell after four weeks of a 2% cream — established (instrument-measured)
  • Whether it beats a base with nothing in it — not established
  • It was the slowest of the three to control oil — established
  • What happens beyond four weeks — not established

If you want more on sebum, read the niacinamide and salicylic acid pages. Both have a wider evidence base than this ingredient.

How to use it

  • The trial used 2% over four weeks.
  • It took the longest of the three arms to work. This is not an ingredient you can judge from a short trial of it.
  • The cream base in that trial was dimethicone and glycerin. The base is part of the result.
  • The arm combining it with EGCG did better than this ingredient alone.
  • Less sebum and less shine are different things. A product that absorbs surface oil reduces shine without reducing how much sebum is produced.

Cautions

  • We have not verified the range and frequency of adverse events reported in the trial. It records only that no problems arose in any of the three arms.
  • There is one human trial.
  • We found no established data on use during pregnancy.
  • Data on oral carnitine does not carry over to applying it.
  • Producing a lot of sebum is not in itself a disease. The oiliness page sets that out.

With a new ingredient, apply a small amount somewhere inconspicuous, such as the inner arm, for two days first. Stop if it reddens or stings.

Related skin concerns

Oiliness and sebum

What this article relies on

  1. PubMed — 미국 국립의학도서관 문헌 데이터베이스National Library of Medicine · Checked on 2026-09-16