The reported study used the dual-task methodology to assess the attention demands associated with high and low contextual interference (CI) practice environments. Two specific issues were addressed. First, is there a difference in the attention demands during random and blocked schedules of practice? Second, what is the time course of any differential attention demands that emerge during random and blocked training? In order to address these questions two specific temporal loci were probed during practice: a pre-response interval and the inter-trial interval. It was assumed that the pre-response interval contained the reconstructive activity that is central to the reconstruction position. In contrast, the inter-trial interval has been interpreted in previous work to be the interval in which critical intra- and inter-item processing is performed during random practice. The data revealed a typical CI effect for the primary key-pressing task. Specifically, blocked-practice participants displayed superior performance during training but performed less well than the random-practice individuals at the time of retention. The poorer acquisition performance of the random-practice participants was associated with higher cognitive demand during both the pre-response and the inter-trial intervals than that of individuals assigned to blocked practice. The greater attention demands for random-practice individuals are discussed with respect to processes that might occur in both the pre-response and the inter-trial intervals.