A stochastic production frontier approach was used to estimate input distance functions for U.S. grass-fed beef (GFB) production. Average technical efficiencies of 0.84 and 0.79 were found for U.S. GFB whole farms and enterprises, respectively. Producer education level, experience, farm size, annual net farm income from the GFB operation, annual net household income from off-farm sources, and regional differences are the efficiency drivers of U.S. GFB farms. Increasing returns to scale were found for U.S. GFB farms. Our results suggest that U.S. GFB farms can be scale efficient if the optimal size of the operation is greater than approximately 100 GFB animals.