Because methamphetamine exposure to experimental animals can damage brain dopamine neurones, we examined whether hospital patients diagnosed with methamphetamine-related disorders might have greater risk of subsequent admission with a Parkinson's disease diagnosis. This was a population-based cohort study using all statewide inpatient hospital discharge records from July 1, 1990, through June 30, 2000, in California, USA, in which subjects aged at least 50 years were followed for up to 10 years. Individuals with reported methamphetamine-related conditions (n = 1,863; ICD-9 codes 304.4, 305.7, 969.7, and E854.2) were matched on demographic variables and follow-up time with those with primary appendicitis conditions (n = 9,315). The appendicitis group had a Parkinson's disease incidence rate no different than the rate found among members of a large health maintenance organization in California. Cox regression procedures were used to estimate group differences in the rates of receiving a subsequent inpatient diagnosis of Parkinson's disease (ICD-9 332.0). The methamphetamine group showed increased risk of a subsequent admission with Parkinson's disease compared with that of the matched appendicitis group (adjusted hazard ratio = 2.65, 95% CI, 1.17-5.98, P= 0.019). Study limitations include a population limited to hospital admissions, an uncertainty regarding diagnostic validity of the ICD-9 code 332.0 (Parkinson's disease), and a small number of incident cases with suspected Parkinson's disease. We strongly emphasize the preliminary nature of the findings. Nevertheless, these data, requiring replication, provide some evidence that methamphetamine users might be at greater than normal risk for developing Parkinson's disease.