As rheumatoid arthritis (RA) patients have been shown to have impaired plasma fibrinolysis and fish oil has been suggested to be useful for RA, this study investigated the effects of fish oil on fibrinolytic parameters in patients with RA. Forty-five RA patients were randomised to receive either fish oil (1.7 gm eicosapentaenoic acid and 1.1 gm docosahexaenoic acid/day) or placebo treatment for at least 6 months. Plasma levels of fibrinogen, tissue-plasminogen activator (t-PA) and plasminogen activator inhibitor (PAI) activity were measured at 3-month intervals. In the fish oil treatment group, plasma levels of fibrinogen and t-PA activity were reduced at 6 months when compared with baseline [fibrinogen: 3.2 (2.85-3.53) milligrams vs 3.89 (3.56-4.22) milligrams, mean (95% confidence intervals for mean), p < 0.02; t-PA activity 1.4 (1.01-1.78) units/ml vs 1.94 (1.55-2.33) units/ml, p < 0.01]. No significant changes in plasma PAI activity were seen during the treatment period in these patients. Placebo treatment did not significantly alter the plasma levels of fibrinogen or t-PA and PAI activity. In conclusion, fish oil supplementation does not appear to produce an improvement in plasma fibrinolysis. Indeed, plasma fibrinogen levels and t-PA activity were reduced. This could be due to an effect of fish oil on acute phase protein production. Alternatively, as t-PA is produced on an "on demand" basis, its reduction may be related to the lowering of fibrinogen levels following fish oil therapy.