This paper develops and explains an empirical method of using cross-sectional data to determine cost functions for forestry operations. Primary logging firm data, basic production economics theory, and regression analysis were used to estimate aggregate cost functions by harvest system for southern United States pulpwood harvesting operations. Estimation of aggregate total costs first, followed by mathematical derivation of average costs, proved superior to direct estimation of average costs. Exponential functions were best at estimating the cost relationships observed in the empirical pulpwood harvesting data.