Recovery planning for threatened and endangered steelhead requires measurable, objective criteria for determining an acceptably low risk of extinction. Here we propose viability criteria for two levels of biological organization: individual populations, and groups of populations within the South-Central/Southern California Coast Steelhead Recovery Planning Domain. For populations, we adapt criteria commonly used by the IUCN (The World Conservation Union) for identifying at-risk species. For groups of populations we implement a diversity-based “representation and redundancy rule,” in which diversity includes both life-history diversity and biogeographic groupings of populations. The resulting criteria have the potential for straightforward assessment of the risks posed by evolutionary, demographic, environmental, and catastrophic factors; and are designed to use data that are readily collected. However, our prescriptive approach led to one criterion whose threshold could not yet be specified due to inadequate data, and others in which the simplicity of the criteria may render them inefficient for populations with stable run sizes or stable life-history polymorphisms. Both of these problems could likely be solved by directed programs of research and monitoring aimed at developing more efficient (but equally risk-averse) “performance-based criteria.” Of particular utility would be data on the natural fluctuations of populations, research into the stabilizing influence of life-history polymorphisms, and research on the implications of drought, wildfires, and fluvial sediment regimes. Research on estuarine habitat could also yield useful information on the generality and reliability of its role as nursery habitat. Currently, risk assessment at the population level is not possible due to data deficiency, highlighting the need to implement a comprehensive effort to monitor run sizes, anadromous fractions, spawner densities and perhaps marine survival. Assessment at the group level indicates a priority for securing inland populations in the southern Coast Ranges and Transverse Ranges, and a need to maintain not just the fluvial-anadromous life-history form, but also lagoon-anadromous and freshwater-resident forms in each population.