Thirty patients (18 women, 12 men) with permanent or paroxysmal atrial fibrillation were treated with a new antiplatelet drug, ticlopidine, in order to study platelet aggregation behaviour, to see whether the drug prevents thromboembolisms and to observe side-effects over a period of 6 months. A further comparative study of the effects of ticlopidine and dipyridamole + aspirin on platelet aggregation was carried out in 20 patients. All appropriate haematological parameters were tested every 3 months, while platelet aggregation curves with ADP were examined also after 15 days. At the end of the period an echocardiogram was performed to check for any sign of thrombosis. The reduction in the aggregation curves was statistically significant for all the ADP stimuli, except at low doses. In the comparison with dipyridamole + aspirin, ticlopidine gave better results; with the former there was no significant reduction in platelet aggregation. A more significant reduction was seen in patients who had showed hyperaggregation at the outset. Bleeding time was increased and platelet adhesivity was reduced. During the 6-month period a slight reduction in white blood cells and a slight increase in creatinine were observed, both remaining within the normal range. Some 33.3% of the patients experienced side-effects. No embolic event or thrombosis in the left atrium was seen. Ticlopidine seems to be an effective antiplatelet drug, especially for patients with hyperaggregation.