In two of the three experiments that are reported subjects were trained in either a blocked or a random practice schedule and allowed to self-select the amount of time they used to plan an upcoming movement. According to the reconstruction hypothesis (Lee & Magill, 1985) random practice participants engage greater movement planning processes (i.e. action plan re-construction) during acquisition than do their blocked practice counterparts. It was predicted that such reconstructive activity would be manifest as greater study time being used by random practice subjects as they readied themselves for an imminent response. Data from Experiments 1 and 2 supported this study time prediction while also revealing the typical retention benefit associated with random practice. Interestingly, the acquisition benefit often apparent with blocked practice was not evident when the random practice participants were given sufficient time to plan the upcoming response. These conclusions were supported by data from Experiment 3 that involved a direct manipulation of the amount of study time afforded random and blocked practice participants. Taken together, these data are consistent with the suggestion from the reconstruction hypothesis that more time is necessary for intratrial planning processes during random practice formats. Furthermore, acquisition differences that are commonly observed between high and low contextual interference practice conditions may, at least in part, be associated with the completion of the extra planning processes entertained by random practice participants as the movement is being executed.