Stream temperature and channel disturbance are two potentially important controls on the distribution and persistence of fish populations. Temperature regulates primary physiological processes that constrain the demographic response of populations to their environments. Ultimately temperature may be a first order determinant of the patterns of potential habitat and occurrence for many species. Stream temperature can be estimated from locally derived empirical relationships with elevation or based on detailed energy balances and thus used to model the distribution of potential habitats for fishes across whole landscapes. The role of disturbance is more hypothetical. Metapopulation theory proposes that environmental variation may have an important influence on the dynamics of populations. Disturbances may depress or even eliminate local populations, but a regional population may persist because other populations are not affected. Demographic support or recolonization may occur through dispersal among populations. Clearly the scale of disturbance and population structure can be important. If the characteristic size of disturbances is larger than the extent of a local population, then adjacent populations may decline simultaneously and metapopulation structure will offer little benefit. Conversely, if the characteristic size is smaller the benefit of structure could be important. In this paper we examine the spatial scale of large disturbances in the Boise River catchment over the last 50 years. We compare that to the scale of habitat patches for bull trout defined by stream temperature and the patterns of genetic variation detected by molecular techniques. Implications for species conservation are discussed in the context of climate change (influencing habitat patch size) and fire and fuels management (influencing the scale of disturbance).