This experiment examined the impact of selective postevent discussion of high- and low-elaborative styles on 5- and 6-year-olds' (N = 47) memory for discussed and nondiscussed aspects of a staged event ("Visiting the Pirate"). The event contained both logically and arbitrarily connected scenes. Discussion was spaced over 3 days, and memory was assessed 1 day later. Compared with a no-discussion condition, memory for discussed information was enhanced after high- but not low-elaborative discussion for both logically and arbitrarily connected scenes. For arbitrarily connected scenes, memory for nondiscussed aspects was impaired relative to the no-discussion condition, with the degree of impairment being equal after high- and low-elaborative discussion. In contrast, for logically connected scenes, memory for nondiscussed information was not impaired after discussion of either style.