A single bolus dose of LMW heparin at the start of haemodialysis effectively prevents clot formation in the dialyser and bubble trap. However, there are few studies on the appropriate dosage of LMW heparins in haemodialysis. Therefore we examined the relationship between the anticoagulant effect of dalteparin and clinical clotting during haemodialysis. We performed an open, prospective study on the effect of decreasing doses of dalteparin in 12 haemodialysis patients during a total of 84 sessions (4-4.5 h). The normally applied dose of dalteparin in each patient was reduced by 25% for each session down to 50% of initial dose if no clotting was observed. Clinical clotting (grade 1-4) was evaluated by visual inspection after blood draining of the air trap every hour and by inspection of the dialyser after each session and compared to corresponding values for anti-FXa activity and dialysis time. Blood flow and ultrafiltration rate were kept within narrow limits throughout the study. No episodes of grade 4 clotting occurred, and no session was interrupted. Eighteen episodes of grade 3 clinical clotting (11%) were observed in patients without warfarin treatment, none with an anti-FXa activity >0.43 IU/ml. Oral warfarin treatment reduced the clinical clotting, and only one grade 3 episode was observed in patients on warfarin therapy. Anti-FXa activity and haemodialysis time were the only factors independently correlated to clotting in a logistic regression model. An anti-FXa activity above 0.4 IU/ml after 4 h of dialysis inhibits significant clotting during haemodialysis. A bolus dose of dalteparin of 70 IU/kg usually seems appropriate, but may be reduced in patients on warfarin treatment. Dialysis time is an independent risk factor for clinical clotting.