Comparisons of Surber sampling of natural substrates and multiplate artificial substrates were made in Trout Creek, a mountain stream in northwest Colorado. Multiplate samplers were placed in the stream and allowed to be colonized for five weeks. Surber samples were taken when multiplates were collected. Multiplate samples generally provided statistically lower density and biomass estimates and number of species than the Surber samples although species diversity was similar with both methods. The seasonal pattern in abundance was dissimilar for the two methods. Greatest density from Surber sampling was found in fall whereas multiplate artificial substrates had their lowest density in fall. Coefficients of variation for biomass values were consistently lower on the multiplates than for the Surber samples but varied considerably for density and number of species. It was concluded that multiplate samplers were not as practical as natural substrate sampling in Rocky Mountain streams.