In this study I try to understand how the approval rating of the Prime Minister affects the probability of his resignation. My analysis relies on the statistical technique of duration modeling, which can be used to explore the effects of independent variables on the occurrence and timing of an event of interest. By applying a discrete-time model on the data set of the postwar Japanese Prime Ministers, I show that the probability of the Prime Minister resigning not only increases over his term in office, but also depends on his ability to gain public support.