Visual perspective-taking is the ability to perceive the world from another person’s perspective, and research on visual perspective-taking ability in children with autism spectrum conditions yielded inconsistent results. To solve a visual perspective-taking task, people can mentally rotate themselves to another person’s location (embodied self-rotation strategy) or else rotate the object toward themselves (object-based mental rotation strategy). Previous interventions for autistic individuals have mainly focused on embodied self-rotation strategy, the visual perspective-taking mechanism in neurotypicals. This study examined the effects of both embodied self-rotation and object-based mental rotation strategies in improving visual perspective-taking performance in 34 autistic children and 34 ability-matched neurotypical children. All children completed three tasks: one measuring baseline visual perspective-taking performance, while the other two measuring visual perspective-taking performance after instructions in embodied self-rotation and object-based mental rotation strategies. The results indicated that autistic children had difficulty in baseline visual perspective-taking tasks. After a brief strategy learning period, autistic children benefited similarly from both embodied self-rotation and object-based mental rotation instructions, suggesting that there are various possible ways to improving autistic children’s perspective taking, and that further interventions for autistic children could consider combining different strategies that better suit their autistic traits. Lay abstract When answering how the same object might appear to others in different locations, people can provide answers by mentally putting themselves into another person’s location using the embodied self-rotation strategy or by rotating the target object toward themselves using the object-based mental rotation strategy. In this study, after learning the embodied self-rotation or object-based mental rotation strategies, autistic children improved their visual perspective-taking performance, which is believed to be impaired or delayed in autistic individuals. We recruited 34 autistic children and an equal number of ability-matched typical children and examined their visual perspective-taking performance at baseline and after learning the embodied self-rotation and object-based mental rotation strategies. As previous visual perspective-taking and other social cognition interventions for autistic individuals have primarily focused on the embodied self-rotation strategy, showing moderate effectiveness and limited generalizability, we explored the effects of both embodied self-rotation and object-based mental rotation strategies for improving perspective-taking performance and discussed their implications in this study. The results showed that autistic children had a lower performance at baseline compared with typical children; however, they were still sensitive to both embodied self-rotation and object-based mental rotation strategies. Unlike typical children, who gained more from the embodied self-rotation strategy, autistic children benefited similarly from the two strategies. This suggests that there are multiple ways to helping autistic children overcome their difficulty in perspective-taking tasks. Future interventions for autistic children could consider combining various strategies that better suit their autistic traits.