The Bureau of Mines is conducting research to develop methods to control fragmentation and improve productivity in surface mine blasting. As part of this research, PRONTO2D, a two-dimensional finite element computer program developed at Sandia National Laboratories, was employed to model the fragmentation produced from RS-50, a single-hole, reduced-scale bench blast in a massive dolomite. The fragment size distribution obtained from the computer simulation was calculated using a submodel that utilizes the crack density and nominal flaw size values computed by PRONTO2D. The computer program successfully modeled the plus 1.5-to minus 9-inch fragmentation observed from the RS-50 blast. The results assume that these fragment sizes were created from stress wave mechanisms only, and that little or no stress wave reflection off the borehole-rock interface occurred. Parameters used for the rock and explosive models in the computer simulation were obtained theoretically or from previously published data. In general, the larger pieces were forecast to form near the free surfaces of the bench and decrease in size towards the borehole with the smallest fragments located almost adjacent to the borehole wall. This result agrees with observations of high-speed films from similar reduced-scale blasts and with small-scale blasting experiments in concrete. The minus 1.5-inch RS-50 fragment distribution was not accurately modeled by the computer simulation. The inability of PRONTO2D to simulate near-borehole fracturing and late failure mechanisms, such as those related to gas pressure effects and initial rock motion, may be responsible for this result. Using the successful computer simulation of the RS-50 test blast as a calibration for the computer model, the relative change in fragmentation from differing burden lengths was studied. PRONTO2D predicted a shift in the fragment distribution to a higher percentage of coarser size fragments as the bench burden was increased, a result consistent with field and laboratory observations. Modification of the stress waves relative to changes in the bench geometry was responsible for the varying predicted fragment distributions.