Mentally simulating a movement is known to share temporal and kinematic characteristics with the execution of the same movement, and this is thought to be reflected in the sharing of neural resources between the two activities. A powerful method of implicitly facilitating such mental simulation (or motor imagery) in individuals is to present them with a picture of a hand and ask them to identify its laterality (i.e. left or right). The mental rotation undertaken in order to complete this hand laterality recognition task (HLRT) provides an effective form of motor imagery, and the task has become an influential tool in clinical and experimental studies. However, performance on the task is modified by numerous factors, and there is a suggestion that the method of response demanded by different versions of the task may have a modulating effect. Here, we compared performance on the HLRT when responding verbally or manually in a group of unimpaired right-handed participants. For manual responses, we also compared performance when participants responded unimanually, using the index and middle fingers of their dominant or non-dominant hand. Performance was poorer for the manual compared to the verbal condition both in terms of accuracy and response time. Furthermore, for manual responses, the requirement to make a response with a specific limb selectively disrupted the ability to recognise an image of the corresponding limb. The disruption is considered to reflect difficulty in concurrently planning two actions with the same limb (manual response and mental rotation). Implications for the interpretation of existing and future studies are discussed.