Using case-control methodology, we assessed prevalence and duration of exposure to putative risk factors for rapid progression of Parkinson's disease (PD) in patients not taking levodopa or direct dopamine agonists. We identified 31 patients termed "rapidly progressive" who were stage I or II (Hoehn and Yahr) on their first visit to our center and who progressed to stage III during the study period; we pair-matched this group with 31 "slowly progressive" patients who had the same symptom duration and the same Hoehn and Yahr stage at study entry, but whose parkinsonism did not progress to stage III during the study. Only age of PD onset was associated with different rates of PD progression; older patients at PD onset progressed more rapidly than younger patients.