Global climate change is apparent within the Arctic and the south-western deserts of North America, with record drought in the latter reflected within 640 000 km2 of the Colorado River Basin. To discern the manner by which natural and anthropogenic drivers have compressed Basin-wide fish biodiversity, and to establish a baseline for future climate effects, the Stream Hierarchy Model (SHM) was employed to juxtapose fluvial topography against molecular diversities of 1092 Bluehead Sucker (Catostomus discobolus). MtDNA revealed three geomorphically defined evolutionarily significant units (ESUs): Bonneville Basin, upper Little Colorado River and the remaining Colorado River Basin. Microsatellite analyses (16 loci) reinforced distinctiveness of the Bonneville Basin and upper Little Colorado River, but subdivided the Colorado River Basin into seven management units (MUs). One represents a cline of three admixed gene pools comprising the mainstem and its lower-gradient tributaries. Six others are not only distinct genetically but also demographically (i.e. migrants/generation <9.7%). Two of these (i.e. Grand Canyon and Canyon de Chelly) are defined by geomorphology, two others (i.e. Fremont-Muddy and San Raphael rivers) are isolated by sharp declivities as they drop precipitously from the west slope into the mainstem Colorado/Green rivers, another represents an isolated impoundment (i.e. Ringdahl Reservoir), while the last corresponds to a recognized subspecies (i.e. Zuni River, NM). Historical legacies of endemic fishes (ESUs) and their evolutionary potential (MUs) are clearly represented in our data, yet their arbiter will be the unrelenting natural and anthropogenic water depletions that will precipitate yet another conservation conflict within this unique but arid region.