I will discuss recent work with K. Becker and G. Guo computing alpha' corrections to type II D-brane actions. I will explain why these corrections are important and will show how T-duality implies the presence of some novel couplings. Disc amplitudes can be used to check these predictions and go beyond them, and I will review some of the subtleties involved in computing the amplitudes and recovering an action. Finally, I will present the complete set of four-derivative couplings that we have computed so far.