- An inextricable link exists between dams and human development in the Pacific Northwest, but they can fragment rivers and reduce genetic connectivity for freshwater and anadromous fishes. Since the early-twentieth century, a series of hydropower and irrigation dams on the mainstem Klamath River, California, has fragmented migratory corridors and eliminated access to spawning habitat for anadromous rainbow trout (Oncorhynchus mykiss), known as steelhead, in the upper reaches of the Klamath Basin extending into southern Oregon. Here, we characterize neutral and adaptive genetic diversity across collections of anadromous and resident coastal rainbow trout (O. mykiss irideus) and Klamath redband trout (O. mykiss newberii) prior to a large-scale dam removal effort. We used a combination of population- and individual-based analyses to describe population structure, test for temporal changes in genetic diversity, and characterize diversity at loci associated with anadromy versus residency and timing of freshwater entry and arrival time on spawning grounds. We found broad structure based on presumed genetic lineage (i.e., coastal rainbow or Klamath redband) and presumed life history (i.e., anadromous, adfluvial, fluvial and freshwater resident) and identified previously unknown substructure among presumed adfluvial redband trout within major river systems in the Upper Klamath Basin. We identified isolated collections of coastal rainbow trout above the current dam sites in Upper Klamath Lake tributaries that appear distinct from potential historical hatchery sources on the mainstem Klamath River and one stock that is currently released in the Upper Klamath Basin. Additionally, we identified likely source populations for presumed adfluvial redband trout sampled in Upper Klamath Lake prior to beginning their spawning migrations. We found moderate differences in genetic differentiation between samples collected greater than 10 years apart in the Lower and Upper Klamath Basin, but not in areas impounded by dams in the mainstem Klamath River. This indicates that genetic diversity is generally stable in areas impacted by the dams. However, at sites with no obvious barriers to migration, low to moderate gene flow and genetic drift may be responsible for the observed genetic differentiation between the same resident populations sampled greater than 10 years apart. At loci associated with anadromy versus residency, we found that across most collections of fish in the Upper Klamath Basin above the current dam sites, collections were nearly fixed for the resident-associated alleles. Notable exceptions were tributaries of the Wood and Sprague River, where several heterozygotes were observed, and a collection in the Upper Williamson River upstream of Klamath Marsh where a large proportion of the fish was heterozygous or homozygous anadromous. In contrast, coastal rainbow trout between dams harbored the anadromous-associated alleles at frequencies comparable to known anadromous collections downstream in the Trinity River. At one locus associated with timing of freshwater entry and arrival time on spawning grounds, we observed a mixture of early-returning, heterozygous, and late-returning genotypes in the Lower Klamath Basin and in the Klamath River between Iron Gate Dam and Link River Dam. In the Upper Klamath Basin, most fish had the early-returning genotype. However, we found several collections in the Sprague River and one collection in the Great Basin with a large proportion of heterozygous fish. At one site in the Upper Klamath Basin with two different run types of presumed adfluvial redband trout, we found that there were no significant differences at all loci associated with timing of freshwater entry and arrival time on spawning grounds.