Supplement sales are enormous and the evidence for most of them in people who eat adequately is weak or absent. A few cases are different and worth knowing.
Where there is reasonable evidence
Vitamin D for people with limited sun exposure, which in high-latitude countries during winter is most of the population. Deficiency is common and testing is straightforward.
Folic acid before and during early pregnancy, where the evidence for preventing neural tube defects is strong and the recommendation is universal.
Vitamin B12 for people eating no animal products, where deficiency develops over years and the consequences are serious.
Iron for people with diagnosed deficiency, which is common in menstruating women. Not otherwise, since excess iron is harmful.
Where the evidence is weak
Multivitamins in people eating a reasonable diet. Large trials have generally not found benefit for the outcomes people take them for.
Most single-nutrient supplements taken without a deficiency. Correcting a deficiency helps; adding more to an adequate level generally does not.
Antioxidant supplements, where some trials found harm rather than benefit.
The pattern
Supplements work when they correct a deficiency. They do not work as insurance against a diet that is otherwise fine.
How to decide
Testing rather than guessing. A blood test for vitamin D, B12 or iron costs little and answers the question, where taking a supplement indefinitely answers nothing.
The caveat
General information. Anyone considering supplements, particularly with a health condition or other medications, should ask a doctor or pharmacist.