Three putative risk factors to the pathogenesis of end-of-dose-related fluctuations--severity of Parkinson's disease (PD), duration of PD, and duration of levodopa therapy--were evaluated in 39 patients with PD by measuring response duration and response magnitude to a single levodopa dose. After discontinuing levodopa therapy for 12 hours, the motor status of each patient was assessed before a single dose of 100 mg of levodopa plus 25 mg of decarboxylase inhibitor was given and at five 45-minute intervals thereafter. Multiple regression analysis showed that the response duration was best predicted by the severity of PD. The response magnitude was significantly correlated both to the duration of PD and to the duration of levodopa therapy. It is argued, however, that the response magnitude was predicted by the duration of levodopa therapy. Our results are in accordance with those of other recent studies and indicate that variances in levodopa response duration and in levodopa response magnitude are caused by their own differential risk factor.