Three experiments examined the functional relationship between the frequency of prospective responding and the prospective interference effect within the context of a task switching paradigm. Prospective responding was less frequent across the experiments when prospective cues appeared in switch blocks than when they appeared in pure blocks. The magnitude of the prospective interference effect for response time (RT) was similar for pure and switch blocks when exogenous task cuing was used, and was greater for switch blocks than for pure blocks when endogenous task cuing was used. These data reveal a dissociation between the effect of task switching on the frequency of prospective responding and the prospective interference effect, and indicate that the functional relationship between these two measures is dependent on task demands.