The problem of scheduling input applications can be examined by extending conventional production function analysis. Using appropriately designed agricultural experiments, it is possible to estimate production function parameters with alternative specifications for input timing (and amount). A study of nitrogen applications to rice is employed to illustrate scheduling via production functions. Alternative specifications and functional forms are simultaneously examined to determine the sensitivity of economic results to these factors. Sensitivity is found to be high, and this finding is hypothesized to be critical for other approaches to input scheduling as well.