Experimental therapies for Parkinson's disease (PD) are commonly validated in unilateral animal models using simple tests of motor asymmetry such as rotation, stepping and cylinder tests. However, the human disorder is considerably more complex than this, and alternative tests that permit a more complete evaluation of the efficacy and mechanism of action of novel treatments are needed. In this study, an operant task that assesses the selection, initiation and execution of lateralized movements was used to investigate the effects of embryonic dopamine cell transplants in the unilateral medial forebrain bundle (MFB) lesion model of PD. Lesioned Lister Hooded rats had a pronounced contralateral selection and initiation deficit, as well as an impairment in execution of movements bilaterally. They also attempted fewer trials and made more procedural errors than unlesioned rats. Transplantation of fetal dopaminergic neurons to the striatum led to a marked improvement in specific parameters and a more modest improvement in others. The graft improved the contralateral selection deficit and the execution of movements bilaterally, but had no effect on the initiation of contralateral movements. Transplanted rats also attempted more trials and made fewer errors. In contrast, the more commonly used stepping and cylinder tests revealed no functional effect of the graft. This data suggests that this operant task may be a powerful tool for validating and elucidating the mechanism of action of experimental brain repair therapies prior to entering the clinic.