Much research has tested techniques to improve children’s reporting of episodes from a repeated event by interviewing children after they have experienced multiple episodes of a scripted event. However, these studies have not considered any effects of the similarity shared between event episodes on children’s reports. In the current study, 5- to 9-year-olds experienced four episodes of a scripted repeated event that shared a high (n = 76) or low (n = 76) degree of similarity, and were subsequently interviewed about individual episodes. The proportional amount and accuracy of children’s reported details were tallied. Children reported proportionally more details and more script deviations after experiencing the high, compared to low, similarity event. Conversely, children were more accurate in their episodic reports when they experienced the low, compared to high, similarity event. The current findings have implications for the generalisability and comparability of past results across laboratory studies.