Parkinson’s disease (PD) patients have been found to exhibit dopamine-dependent learning impairments. However, performance on learning tasks depends on reinforcement learning (RL) and working memory (WM). It remains unclear whether learning impairments in PD are attributed to RL or WM. We hypothesize that dopaminergic medication improves RL and WM and that PD patients demonstrate dopamine-dependent RL deficits compared to healthy controls after WM deficits are controlled. A total of 191 PD patients and 50 healthy controls were recruited. PD patients rated their subjective dopaminergic medication states on self-report scales, and all participants completed the now well-established Reinforcement Learning Working Memory task. We will adopt computational modeling approaches to disentangle the contributions of RL and WM and test the effects of medication states on RL after controlling for WM. The study has implications for a better understanding of how dopaminergic medication influences cognitive performance, thus providing insights for individualized PD treatments.