I examined how land use influenced the distribution of Smallmouth Bass Micropterus dolomieu in channel units (discrete morphological features—e.g., pools) of streams in the Midwestern USA. Stream segments (n = 36), from four clusters of different soil and runoff conditions, were identified that had the highest percent of forest (n = 12), pasture (n = 12), and urban land use (n = 12) within each cluster. Channel units within each stream were delineated and independently sampled once using multiple gears in summer 2006. Data were analyzed using a generalized linear mixed model procedure with a binomial distribution and odds ratio statistics. Land use and channel unit were strong predictors of age-0, age-1, and age->1 Smallmouth Bass presence. Each age-class was more likely to be present in streams within watersheds dominated by forest land use than in those with pasture or urban land uses. The interaction between land use and channel unit was not significant in any of the models, indicating channel unit use by Smallmouth Bass did not depend on watershed land use. Each of the three age-classes was more likely to use pools than other channel units. However, streams with high densities of Smallmouth Bass age >1 had lower proportions of pools suggesting a variety of channel units is important even though habitat needs exist at the channel-unit scale. Management may benefit from future research addressing the significance of channel-unit quality as a possible mechanism for how land use impacts Smallmouth Bass populations. Further, management efforts aimed at improving stream habitat would likely be more beneficial if focused at the stream segment or landscape scale, where a variety of quality habitats might be supported.