Reaction time (RT) and movement time (MT) are reported to be delayed in Parkinson's disease (PD), but their clinical utility and relationship with clinical findings is still uncertain. We investigated RT and MT in 22 PD patients at baseline conditions and following acute oral trials of levodopa and biperiden, an anticholinergic drug. At baseline conditions, RT and MT of PD patients were abnormally delayed compared with those of 16 normal control subjects. Both RT and MT were longer in more severely affected patients compared with the mild PD patients; in the mild PD patients with asymmetrical signs both responses were longer on the more affected side. Bradykinesia was the clinical symptom that best correlated with the objective measurements, with a stronger correlation for MT than for RT. The oral administration of levodopa significantly improved both the responses, whereas biperiden was ineffective. The magnitude of RT and MT improvement after levodopa differed; MT improvement was related to PD severity, whereas RT improvement was not. These results suggest that MT, rather than RT, is an objective, simple, and reliable tool to evaluate bradykinesia and its levodopa-induced modifications in PD.