In a case-control study, we compared the past dietary habits of 342 Parkinson's disease (PD) patients recruited from nine German clinics with those of 342 controls from the same neighborhood or region. Data were gathered with a structured interview and a self-administered food-frequency questionnaire, and analyzed using multivariate conditional logistic regression to control for educational status and cigarette smoking. There was no significant difference between cases and controls in the consumption of fruits and vegetables, although there was a negative trend for the consumption of raw vegetables. Controls reported a higher potato consumption than patients (OR = 0.43, 95% confidence interval [CI]: 0.24-0.74, highest versus lowest quartile). Patients reported eating significantly larger quantities of sweet foods as well as having more snacks than controls. This may, however, be the result of an illness-related change in dietary habits leading to a selective recall effect, since sweet foods may enhance the transport of L-dopa across the blood-brain barrier. We also found that patients consumed less beer (OR = 0.26, 95% CI: 0.14-0.49) and spirits (OR = 0.56, 95% CI: 0.36-0.86), but not wine, and they consumed less coffee (OR = 0.27, 95% CI: 0.14-0.52, highest versus lowest quartile), but not tea, than controls. This may relate to a possible interaction between dopaminergic activity and the intake of ethanol or caffeine. Significantly more patients than controls reported ever consuming raw meat (OR = 1.78, 95% CI: 1.21- 2.63). These results suggest that the intake of certain foods may be associated with the development of PD.