The aim of this study was to determine how and whether changes in the primary motor cortex (M1) are affected by dual motor task. We further investigated how dual motor task is dependent on task properties measured using transcranial magnetic stimulation (TMS). TMS delivered to left M1 during the dual motor task and motor-evoked potential (MEP) were simultaneously evoked in the right FDI, thenar, FCR and ECR muscles. In experiment 1, subjects were asked to simultaneously walk on a treadmill and perform finger prehension. The gait conditions were employed 30, 50 and 80% of maximum walking speed (gait 30%, gait 50% and gait 80%). Conditions for finger prehension while following the visual tracking task varied with force outputs of 5 and 25% of maximum voluntary contraction (MVC). In experiment 2, the subjects were asked to perform optimal walking synchronized with the finger prehension task with an optimal walking rhythm (2-Hz dual motor task), as well as optimal walking desynchronized with the finger prehension task (0.7-Hz dual motor task). In experiment 1, MEPs were markedly decreased under the gait 50% condition compared with those under the gait 30 and 80% conditions at 5% MVC. In experiment 2, MEPs were markedly decreased with the 2-Hz dual motor task compared with those with the 0.7-Hz dual motor task. Our results suggest that the excitability changes in M1 during the dual motor task were dependent on changes in the gait speed, precision of prehension task and temporal movement.