This study examined the impact of response repetition as an error-correction strategy in teaching subtraction facts to three students with learning difficulties. Written response repetition (WRR) and oral response repetition (ORR) were compared using an alternating treatments design nested in a multiple baseline design across participants. Introduction of the error-correction procedures resulted in increases in subtraction facts mastered and correct responding relative to baseline for all participants. All participants mastered more facts in the WRR condition, with varying levels of differentiation. Two of three participants demonstrated higher levels of correct responding in the WRR condition. Results are discussed in terms of consistency with other studies of response repetition and implications for school-based personnel developing and delivering intensive academic interventions.