There is conflicting evidence on the benefits of foods rich in vitamin E (α-tocopherol), n-3 polyunsaturated fatty acids (PUFA), and their pharmacological substitutes. We investigated the effects of these substances as supplements in patients who had myocardial infarction. From October, 1993, to September, 1995, 11 324 patients surviving recent (≤3 months) myocardial infarction were randomly assigned supplements of n-3 PUFA (1 g daily, n=2836), vitamin E (300 mg daily, n=2830), both (n=2830), or none (control, n=2828) for 3·5 years. The primary combined efficacy endpoint was death, non-fatal myocardial infarction, and stroke. Intention-to-treat analyses were done according to a factorial design (two-way) and by treatment group (four-way). Treatment with n-3 PUFA, but not vitamin E, significantly lowered the risk of the primary endpoint (relativerisk decrease 10% 95% CI 1–18 by two-way analysis, 15% 2–26 by four-way analysis). Benefit was attributable to a decrease in the risk of death (14% 3–24 two-way, 20% 6–33 four-way) and cardiovascular death (17% 3–29 two-way, 30% 13–44 four-way). The effect of the combined treatment was similar to that for n-3 PUFA for the primary endpoint (14% 1–26) and for fatal events (20% 5–33). Dietary supplementation with n-3 PUFA led to a clinically important and satistically significant benefit. Vitamin E had no benefit. Its effects on fatal cardiovascular events require further exploration.