Several studies have reported an apparent protective effect of cigarette smoking for the risk of idiopathic Parkinson's disease (IPD). These observations are supported by neurochemical studies demonstrating enhancement of central dopaminergic neurotransmission by nicotine. We studied the prevalence and severity of neuroleptic-induced parkinsonism (NIP) in relation to cigarette smoking in a homogeneous sample of 130 psychiatric inpatients receiving long-term neuroleptic treatment. Despite the fact that smokers had significantly higher dosage of neuroleptics during the month prior to evaluation and longer exposure to medication, they presented with significantly less prevalence and severity of NIP than nonsmokers. These findings suggest that the inverse association between smoking and IPD may apply to NIP.