The risk factors for Parkinson's disease (PD) are not well established. We therefore examined the prediction of various lifestyle factors on the incidence of PD in a cohort drawn from the Finnish Mobile Clinic Health Examination Survey, conducted in 1973-1976. The study population comprised 6,715 men and women aged 50-79 years and free of PD at the baseline. All of the subjects completed a baseline health examination (including height and weight measurements) and a questionnaire providing information on leisure-time physical activity, smoking, and alcohol consumption. During a 22-year follow-up, 101 incident cases of PD occurred. The statistical analyses were based on Cox's model including age, sex, education, community density, occupation, coffee consumption, body mass index (BMI), leisure-time physical activity, smoking and alcohol consumption as independent variables. At first, BMI was not associated with PD risk, but after exclusion of the first 15 years of follow-up, an elevated risk appeared at higher BMI levels (P for trend 0.02). Furthermore, subjects with heavy leisure-time physical activity had a lower PD risk than those with no activity [relative risk (RR) 0.27, 95 % confidence interval (CI) 0.08-0.90]. In variance with findings for other chronic diseases, current smokers had a lower PD risk than those who had never smoked (RR 0.23, 95 % CI 0.08-0.67), and individuals with moderate alcohol intake (at the level of <5 g/day) had an elevated PD risk compared to non-drinkers. The results support the hypothesis that lifestyle factors predict the occurrence of Parkinson's disease, but more research is needed.