Technical progress in U.S. agriculture is evaluated using a new measure of productivity growth, flexible technical change. This measure allows for nonconstant returns to scale, market structures other than perfect competition, and time-varying coefficients. An integral part of the procedure is the estimation of the production function by Flexible Least Squares. Flexible technical change results are compared with two traditional measures of productivity growth and found to be more stable and more precise in a statistical sense. The results suggest that previous studies which employed total factor productivity measures may have overstated the impact of technology in agriculture.