In 753 patients with acute myocardial infarction, use of fish oils (FO, n = 242) before onset of infarction seemed to reduce infarct size as estimated from peak creatine kinase (CKmax) and lactate dehydrogenase (LDmax) activities. The study had an observational exposed/nonexposed design, and both crude and adjusted effects were looked for. CRUDE EFFECTS: In the restricted cohort of patients not receiving thrombolytic treatment (n = 411), FO reduced CKmax from 879 to 759 U/l (2 p = 0.030) and LDmax from 870 to 768 U/l (2 p = 0.011), respectively. More of these patients in the lowest enzyme quartiles used FO, p for linear trend was for CKmax 0.008 and for LDmax 0.06, respectively. ADJUSTED EFFECTS: In patients not receiving thrombolytic treatment, FO reduced CKmax (2 p = 0.007) and LDmax (2 p = 0.005), but in patients receiving such treatment, CKmax and LDmax values increased, 2 p being 0.036 and 0.097, respectively. In patients not receiving thrombolysis, FO increased the incidence of small infarcts (the 25% quartile), odds ratio for CKmax was 1.82 (2 p = 0.018) and for LDmax 1.66 (2 p = 0.048), respectively. The results indicate that FO may reduce infarct size and the incidence of large infarcts. In addition, FO seems to enhance the effect of thrombolysis.