Motor performance is influenced by movements that were performed shortly prior. For example, reaction times (RTs) for successive movements are reduced when executed with the same effector, even if the specifics of the consecutive movements differ. These findings have been taken to support the notion that repetition effects in motor planning reflect the involvement of effector-specific motor plans. However, previous studies have confounded motor and visual aspects of repetition: movements have typically been instructed via visual cues, and movement repetition, therefore, implied repeating also the visual cue, so that the latter may be (at least partly) responsible for the observed RT effects. In the present study, participants performed two movements in succession, a prime and a probe action, either with their left or right hand and in one of two directions, inward or outward relative to the body midline. We used different cues for prime and probe actions, so that movement repetitions did not involve repetition of the visual cue. Participants initiated successive same-limb movements faster than different-limb movements, but this RT advantage was smaller than observed in previous work. Moreover, repeating movement direction also led to a decrease in RT, though only in combination with hand repetition. Whereas these findings imply that visual cue repetition can contribute to accelerated RTs in movement repetition, they confirm that the recent motor history affects motor planning. Furthermore, they support the idea of a hierarchical framework of motor planning in which effector selection precedes specification of motor parameters.