This article presents a technique for nonparametrically estimating continuous-time diffusion processes that are observed at discrete intervals. We illustrate the methodology by using daily three and six month Treasury Bill data, from January 1965 to July 1995, to estimate the drift and diffusion of the short rate, and the market price of interest rate risk. While the estimated diffusion is similar to that estimated by Chan, Karolyi, Longstaff, and Sanders (1992), there is evidence of substantial non-linearity in the drift. This is close to zero for low and medium interest rates, but mean reversion increases sharply at higher interest rates.