Studies investigating the assessment of depression in Parkinson's disease (PD) are limited. We examined the concurrent validity and the internal consistency of the Hamilton Depression Inventory (HDI) and compared it to the Hamilton and Geriatric Depression Scales. PD patients (n = 79) were recruited from neurology clinics. Diagnosis of depressive disorder was made according to DSM-IV criteria. Receiver operating characteristic curves were used to calculate sensitivity, specificity, and positive and negative predictive values. The HDI exhibited an optimal cutoff for discriminating between depressed and nondepressed PD patients of 13.5/14.0 and is a valid instrument to use in the setting of PD.