The objective of this study was to investigate whether eye-hand coupling was preserved or not in PD. We studied predictive saccade performance during hand pointing in six Parkinson's disease (PD) patients with asymmetrical motor signs compared to nine age-matched healthy subjects. The motor responses (saccades and hand pointing) were elicited under open loop conditions (without vision of the hand), by a visual target stepping at a predictable location (10 degrees right and left from the centre) and time. The subjects had to simultaneously move the eyes and point with the finger to the visual target alternating at one of three fixed frequencies (0.25, 0.5 and 1 Hz), for 30 cycles. This task was performed in two sessions balanced over the subjects: one session of ocular saccades only and another session of combined ocular saccades and manual pointing. In the PD group, motor performance was perturbed particularly in terms of increased latencies of hand movements. Interestingly, during pointing, associated predictive saccade disorders were tightly related to the defects of the pointing hand. Indeed, with respect to the latency of predictive saccades alone, the predictive saccade latency during hand pointing significantly decreased in the control group and in the PD group when using the non-affected hand. In contrast, for the PD group when using the affected hand, the saccade latency was increased from the latency values of predictive saccades induced without pointing. Moreover, in the control and in the PD groups, the correlation between eye and hand latencies was highly significant, suggesting an intact eye-hand coupling. No saccadic amplitude disorders were found in either condition. These results demonstrate that eye-hand coupling is preserved in PD, as revealed by the possible beneficial or adverse effects on the ocular saccades, respectively, of the less- or more-affected hand motor responses. This eye-hand coupling mechanism likely involves regions other than the nigro-striatal pathways affected in PD.