A diet rich in fish tends to increase the bleeding time both in patients and healthy controls. Another effect of such a diet is a reduction in serum triglyceride levels in the individuals. It is not known whether these two effects are interrelated. In order to rule out effects of selenium in the fish diet, half of a control group consumed bread enriched with selenium while the other half had a normal diet. This combined control group (n = 21) was compared with healthy volunteers (n = 11) eating 250 g of fish daily for six weeks. Blood tests were performed at the start of the study, after three weeks and at the end. Changes in bleeding time were then related to changes in hemostatic variables and serum lipids. An inverse correlation (r = -0.41, p = 0.001) was present in the controls between serum triglyceride levels and bleeding times, but neither of the two variables changed in the course of the study. A negative correlation was observed in the fish eating group, and a change in the variables mentioned occurred in an orderly manner in the course of the study: a prolongation of the bleeding time from the start of the study to the end was only seen in individuals with low triglyceride levels, i.e. less than 1 mmol/L. Selenium in the diet, on the other hand, did not affect any of the observed variables.(ABSTRACT TRUNCATED AT 250 WORDS)