The antisaccadic paradigm can be applied to test the suppression of reflexive saccades and the activation of volitional saccades simultaneously. The impaired frontal cortex has been shown to have difficulty in suppressing reflexive saccade (prosaccade) to make a successful antisaccade. Degraded antisaccade performance can also be observed in patients with Parkinson's disease (PD). The studies of PD based on the prosaccadic and antisaccadic paradigms have shown controversial findings; the latency between patients and age-matched controls could be either with or without significant difference. Even with this inconsistency, our previous study and recent analysis have supported that the latency of both prosaccade and antisaccade increases significantly for patients with PD. The objective of this study is to investigate whether prolonged antisaccade latency is caused by the affected volitional decision process (volition latency) or simply by the delayed initiation of saccade with direction opposite to the cue, by measuring prosaccade and antisaccade latency from the intermingled paradigms. Eleven mildly affected patients with idiopathic PD and eight age-matched normal subjects were tested in this study. As compared to the age-matched control, the results showed that prosaccade, antisaccadic, and volition latency of the patients was significantly elevated (P<0.01). We conclude that antisaccade performance for the PD patients was degraded for both the volition decision process and the initiation of saccade with direction opposite to the cue. Also, volition latency analysis is a more objective method than prosaccade and antisaccade latency analysis, which can be compared among results obtained from different analysis methodologies.