In this study we tested the gateway hypothesis of Broadmann area 10 (BA10). With a functional magnetic resonance imaging (fMRI) protocol we manipulated the saliency--stimulus-oriented (SO) attending--and the memory load--stimulus-independent (SI) attending--during a prospective memory (PM) task. We found a significant main effect of the SO manipulation within the medial BA10 and a significant interaction between SI attending and PM task within the left lateral BA10. Our results give experimental support to the gateway hypothesis.