Estimation of runoff and sediment yield from small, ungaged watersheds is a difficult hydrologic task. Process oriented mathematical models can be an important part in the solution of this task. Models, however, require some information which is directly related to the hydrologic processes occurring on the watershed. Rainfall simulation is an important experimental technique for gathering such information. This technical report contains the results of a pilot study on the utility of using rainfall simulation in southwestern watersheds. Two different simulators were compared on three sites in New Mexico. A small area simulator, 10 square feet, and a large area simulator, approximately 2000 square feet, were operated for a total of 60 plot experiments. The large simulator was modified and operated with variable intensity rainfall and with overland flow without rainfall for some of the experiments. Analysis of the data indicates that the simulators provide similar results for hydrologic processes, and that sediment yields averaged about 2.7 times higher for the small simulator. Addition of the variable intensity and overland flow capabilities to the large simulator have provided a means for isolating and studying those two important processes.