The morphology of mountainous areas is strongly influenced by stream bed incision rates, but most studies of landscape evolution consider erosion at basin scales or larger. The research here attempts to understand the smaller-scale mechanics of erosion on exposed bedrock channels in the conceptual framework of an established saltation-abrasion model by Sklar and Dietrich [2004]. The recirculating flume used in this experiment allows independent control of bed slope, water discharge rate, sediment flux, and sediment grain size – all factors often bundled together in simple models of river incision and typically cross-correlated in natural settings. This study investigates the mechanics of erosion on exposed bedrock channels caused by abrasion of transported particles. Of particular interest are saltating particles, as well as sediment near the threshold between saltation and suspension - sediment vigorously transported but with significant interaction with the bed. The size of these erosive tools are varied over an order of magnitude in mean grain diameter, including a sand of D¬50 = 0.56 mm, and three gravel sizes of 3.39, 4.63, and 5.88 mm. Special consideration was taken to prevent any flow conditions that created a persistent alluvial cover. The erodible concrete substrate is fully exposed at all times during experiments reported here. Rates of erosion into the concrete substrate (a bedrock proxy) were measured by comparing topographic data before and after each experimental run, made possible by a precision laser mounted on a high speed computer-controlled cart. The experimental flume was able to produce flow discharge as high as 75 liters per second, sediment fluxes (of many varieties) up to 215 grams per second, and bed slopes up to 10%. I find a general positive correlation is found between erosion rate and bed slope, shear stress, grain size, and sediment flux.