The aim of this study is to evaluate the acute effect of playing games on executive function and motor ability in Parkinson's disease (PD). Consecutive cases with PD were studied with the Unified Parkinson Disease Rating Scale (UPDRS), Mini-Mental State examination (MMSE), Beck Depression Inventory (BDI), Stroop test, finger tapping and 14-meter walk test. After randomization, patients performed a game of dominoes and were tested before and after experiment being further categorized as control, winners or non-winners. Forty patients, 27 male (67.5%), aged 48 to 84 years (63.2 ± 8.5), Hoehn & Yahr I to III were included. Twenty-eight (70%) presented depressive symptoms (BDI > 10). Groups (Control N = 13; Winners = 14 and Non-winners = 13) were not different regarding age, disease duration, age at onset, BMI, MMSE scores, depressive symptoms, levodopa dose, and previous practice of games. Winners presented significantly better results on executive function (Stroop test, p = 0.002) and on motor activity (Finger tapping, p = 0.01). Non-winners showed a trend of better performance in the 14-meter-walk test. This study shows that the practice of a non-reward game acutely improved memory and motor skills in PD. Our results suggest a role for the reward system in the modulation of the dopaminergic function of the basal ganglia in these patients.