For the first time genetic similarities among chinook salmon and among steelhead trout stocks of the Columbia River were determined using a holistic approach including analysis of life history, biochemical, body shape and meristic characters. We examined between year differences for each of the stock characteristics and we also correlated the habitat characteristics with the wild stock characteristics. The most important principle for managing stocks of Columbia River chinook salmon and steelhead trout is that geographically proximal stocks tend to be like each other. Run timing and similarity of the stream systems should be taken into account when managing stocks. There are similarities in the classifications derived for chinook salmon and steelhead trout. Steelhead trout or chinook salmon tend to be genetically similar to other steelhead or chinook stocks, respectively, that originate from natal streams that are geographically close, regardless of time of freshwater entry. The primary exception Lo this trend is between stocks of spring and fall chinook in the upper Columbia River where fish with the different run timings are dissimilar, though geographically proximate stocks within a run form are generally very similar. Spring chinook stocks have stronger affinities to other spring chinook stocks that originate in the same side of the Cascade Range than to these Spring chinook stock: spawned on the other side of the Cascade Range. Spring chinook from west of the Cascades are more closely related to fall chinook than they are to spring chinook from east of the Cascades. Summer chinook can be divided into two main groups: (1) populations in the upper Columbia River that smolt as subyearlings and fall chinook stocks; and (2) summer chinook stocks from the Salmon River, Idaho, which smolt as yearlings and are similar to spring chinook stocks from Idaho. Fall chinook appear to comprise one large diverse group that is not easily subdivided into smaller subgroups. In general, upriver brights differ from tules by at least one locus. Steelhead stocks can be divided into two main groups: (1) those stocks found east of the Cascades; and (2) those stocks found west of the Cascade Mountains. Steelhead from west of the Cascades are divisable into three subgroups of closely related stocks: (1) a group comprised mainly of wild winter steelhead from the lower Columbia River; (2) Willamette River hatchery and wild winter steelhead; and (3) summer and winter hatchery steelhead stocks from both the lower Columbia and Willamette Rivers. Steelhead from east of the Cascades are separable into three subgroups of closely related stocks: (1) wild summer steelhead; (2) a group comprised mainly of hatchery summer steelhead stocks; and (3) other hatchery and wild steelhead from Idaho. Streams east and west of the Cascades can be differentiated using characters including precipitation, elevation, distance from the mouth of the Columbia, number of frost-free days and minimum annual air temperature. There are significant differences among the stocks of chinook salmon and steelhead trout for each of the meristic and body shape characters. Between year variation does not account for differences among the stocks for the meristic and body shape characters with the exception of pelvic fin ray number in steelhead trout. Characters based on body shape are important for discriminating between the groups of hatchery and wild steelhead stocks. We could not determine whether the basis for the differences were genetic or environmental. The reason for the variation of the characters among stocks is as yet unclear. Neutrality or adaptiveness has not been firmly demonstrated.