While the role of task repetition has received much attention in task-based research, few studies have examined how exact task repetition affects the performance of child second language learners. Also, little is known about the impact of exact task repetition on trade-off effects between linguistic performance areas among child learners. To help fill this gap, we investigated the impact of task repetition on 40 Chinese EFL learners' oral production. The children repeated the same story-telling task three times, and transcripts of their performance were coded for linguistic complexity, accuracy, and fluency. Complexity was expressed in terms of overall complexity and subordination and phrasal complexity. We assessed accuracy with weighted clause ratios and proportion of errors. Fluency was captured by repair and breakdown fluency measures. Wilcoxon signed rank tests revealed positive effects for task repetition on fluency and accuracy. Trade-off effects observed during participants' first performance had decreased by their third retelling. These results support Skehan's (1998) Limited Capacity model and suggest that task repetition is a useful pedagogical tool in instructed child L2 contexts.