This paper studies whether IMF programs and their size affect borrowing costs by comparing bonds issued immediately before the onset of the program with bonds issued immediately after the program. We show that, on average, the approval of the program leads to a 72-basis points reduction in borrowing costs and that program size matters. Our point estimates indicate that when program size increases by one percent of GDP, borrowing costs decrease by 23 basis points. We also show that program size mostly matters for ex-post programs (i.e., those implemented during crises). For precautionary ex-ante programs, there is some evidence that program size attenuates the reduction in borrowing costs. However, this effect is small and in most cases IMF programs still lead to a statistically significant reduction in borrowing costs.