Targeting specific locations within agricultural watersheds for installing vegetative buffers has been advocated as a way to enhance the impact of buffers and buffer programs on stream water quality. Existing models for targeting buffers of Hortonian, or infiltration-excess, runoff are not well developed. The objective was to improve on an existing soil survey–based approach that would provide finer scale resolution, account for variable size of runoff source area to different locations, and compare locations directly on the basis of pollutant load that could be retained by a buffer. The method couples the Soil Survey Geographic database with topographic information provided by a grid digital elevation model in a geographic information system. Simple empirical equations were developed from soil and topographic variables to generate two indexes, one for deposition of sediment and one for infiltration of dissolved pollutants, and the equations were calibrated to the load of sediment and water, respectively, retained by a buffer under reference conditions using the process-based Vegetative Filter Strip Model. The resulting index equations and analytical procedures were demonstrated on a 67 km2 (25.9 mi2) agricultural watershed in northwestern Missouri, where overland runoff contributes to degraded stream water quality. For both indexes, mapped results clearly mimic spatial patterns of water flow convergence into subdrainages, substantiating the importance of size of source area to a given location on capability to intercept pollutants from surface runoff. A method is described for estimating a range of index values that is appropriate for targeting vegetative buffers. The index for sediment retention is robust. However, the index for water (and dissolved pollutant) retention is much less robust because infiltration is very small, compared to inflow volumes, and is relatively insensitive to the magnitude of inflow from source areas. Consequently, an index of inflow volume may be more useful for planning alternative practices for reducing dissolved pollutant loads to streams. The improved indexes provide a better method than previous indexes for targeting vegetative buffers in watersheds where Hortonian runoff causes significant nonpoint pollution.