Almost every ingredient in this category is described the same way on the packaging: brightening, illuminating, clinically proven. The gap between the ones with substantial clinical backing and the ones coasting on a single small study is enormous, and nothing on the front of the bottle tells you which is which.
Here is the honest ranking, and the numbers that matter on the ingredients list.
How most of these actually work
Melanin production runs through an enzyme called tyrosinase. Most pigmentation actives interfere with it — either by inhibiting the enzyme directly, by blocking the transfer of finished pigment into skin cells, or by speeding up how fast pigmented cells shed.
A few work differently, and those are worth knowing about, because they tend to be the ones that help when the standard approach has stalled.
Strong evidence
Hydroquinone (2–4%)
The long-standing reference treatment, and still the benchmark others are measured against. It directly inhibits tyrosinase and is genuinely effective.
The caveats are real. Prolonged uninterrupted use carries a risk of exogenous ochronosis — a blue-black discolouration that is harder to treat than the original problem. It is generally used in cycles rather than indefinitely. Regulatory status varies significantly: it remains available over the counter in the US at 2%, while in the UK and EU it is prescription-only for cosmetic use.
Tranexamic acid (2–5% topical)
The most interesting development in this category in years, and the evidence has held up well. It works through a different route to the tyrosinase inhibitors — interrupting the plasmin pathway, which reduces the signalling that tells melanocytes to ramp up.
That different mechanism is why it often helps in melasma specifically, where the standard actives underperform. It is also well tolerated, which matters in a condition that worsens with irritation.
Azelaic acid (15–20%)
Genuinely underrated. It inhibits tyrosinase, has anti-inflammatory activity, and is selective — it acts more strongly on hyperactive melanocytes than normal ones, so it tends not to lighten surrounding skin.
It is also one of the few options generally considered acceptable during pregnancy, which matters given how often melasma appears then. Below about 10% you are unlikely to see much.
Retinoids (tretinoin, and to a lesser extent retinaldehyde)
Strong evidence, but mostly as a partner rather than a soloist. Retinoids accelerate cell turnover, which sheds pigmented cells faster, and they improve how well other actives penetrate.
Handle with care in melasma — the irritation retinoids can cause is itself a pigmentation trigger. Start low and infrequent.
Reasonable evidence
Cysteamine (5%)
A relative newcomer with better trial data than most things at this stage of adoption, including some head-to-head work against hydroquinone. The practical problem is the smell, which is genuinely unpleasant, and adherence suffers as a result.
Niacinamide (4–5%)
Doesn't touch tyrosinase. It blocks the transfer of finished melanosomes into keratinocytes — pigment still gets made, but less of it arrives where you'd see it.
The effect is modest on its own and it is rarely a solution by itself. But it is cheap, extremely well tolerated, and stacks well with almost everything, which makes it a sensible base layer.
Vitamin C (L-ascorbic acid, 10–20%)
Real but modest effects on pigmentation, plus genuine antioxidant benefit that helps against the light exposure driving the problem in the first place.
The catch is stability. L-ascorbic acid oxidises readily, and an oxidised serum — you can spot it, it turns yellow-brown — is doing nothing. Formulation and packaging matter more here than with almost any other active. Opaque, air-restricted packaging is close to a requirement.
Thin evidence
Kojic acid
Widely used, particularly in Asian formulations, and it does inhibit tyrosinase. But the human trial data is limited and the effect sizes are small compared to the options above. It also has a reasonable rate of contact sensitisation.
Alpha arbutin
A hydroquinone derivative, marketed as a gentler alternative. It does convert to hydroquinone in the skin — but slowly and in small amounts, which is both why it is gentler and why it is weaker. Most of the enthusiasm outstrips the published evidence.
"Brightening" botanical blends
Liquorice root, mulberry, bearberry and similar extracts appear constantly on ingredient lists. Some have plausible mechanisms in laboratory conditions. Very few have meaningful human trial data at the concentrations actually used in consumer products, which are typically far below what the lab work used.
What actually predicts results
Three things, in order:
- Whether you're protecting against light. This outweighs the choice of active. An average serum plus rigorous daily protection beats an excellent serum without it.
- Concentration and formulation, not the ingredient list. An ingredient present at 0.1% for the label claim is marketing. Products that state their concentrations are telling you something; products that don't, usually aren't.
- Whether you keep using it. Most of these need 8–12 weeks before a fair assessment, and stopping means gradual return. The best formula you abandon at week three loses to a mediocre one you actually stick with.
If you're starting from scratch: a tinted broad-spectrum SPF every morning, azelaic acid or tranexamic acid as the active, niacinamide as a well-tolerated base. Add a retinoid later, carefully, once the routine is stable. Give it three months before judging.