The estimative of the maximum surface runoff volume is fundamental to design several types of hydraulic structures as, for example, level terracing systems. The incorrect estimative of the surface runoff volume causes a wrong design of these systems and when overestimated present high installation cost. Considering these aspects, the present study aimed to compare the values of the surface flow maximum volume obtained by the empiric method of the curve number and developed by the USA Soil Conservation Service, with those obtained by application of the methodology developed by Pruski and collaborators, based on hydrology-consecrated principles. The performance of the two methods was analyzed by considering a 10-year return period and the rain pluviometric conditions corresponding to Belo Horizonte and Uberlândia cities. Great differences were evidenced among the values of the surface flow volume obtained by both methods, which were shown more accentuated for soil water intake rates, and the method proposed by Pruski and his collaborators presented behavior according to reality.
soil conservation; terracing; hydrologic models