This study investigated the effects of recasts, a frequently provided type of oral feedback in language teaching, on the production of English question forms in second language (L2) learning. Despite the large number of studies which have examined the effects of recasts, mixed views exist as to their usefulness as a corrective feedback technique. The present study has continued the investigation into whether their frequent use is justified and also if learners actually perceive recasts as being corrective in nature. The focus of this study is to explore developmental effects, both immediately following and delayed, of recasts while considering these two key aspects. The primary aim is to explore delayed benefits of recasts while the secondary aim is to investigate whether or not these learners perceived the feedback as a correct version of their incorrect question form. The study involved 34 English as a second language (ESL) learners from various first language (L1) backgrounds who were enrolled in a general English programme, at an Auckland language school. The participants completed demographic questionnaires and took part in task-based interaction sessions, with the members of the experimental group receiving input containing intensive recasts. The proportion and types of questions produced in the immediate and two delayed post-tests were compared and analysed to consider whether exposure to recasts was beneficial. The main findings suggest that for more advanced question types, interaction with intensive recasts may be more effective than interaction alone, for delayed L2 development, despite only weak support in the literature. On the other hand, similar positive effects on immediate question production are not strongly evident. Therefore, the results in this study add support to the claim that recasts may be beneficial for delayed second language acquisition (SLA), when the linguistic target is questions, even when these recasts do not seem effective immediately following exposure. Furthermore, awareness of negative evidence was indicated, which may have contributed, in part, to the positive delayed findings. Finally, these results suggest that recasts can be an effective pedagogical tool in the communicative L2 classroom when providing linguistic feedback in a way that does not unduly disrupt the flow of the interactive task. Practical suggestions for future research were also identified.