The average size and age of chinook salmon (Oncorhynchus tshawytscha) caught in commercial fisheries along the Pacific Coast of North America have decreased substantially in this century. These declines might be caused in part by changes in size and age at maturity within the stocks contributing to those fisheries. Upriver Brights (Brights), a stock of fall chinook salmon in the Columbia River, are one of those stocks. The purposes of this study were to (1) determine if average size and age at maturity of Brights have declined, (2) gain a better understanding of the factors that may contribute to such declines, and (3) describe potential consequences of these changes. Data from in-river fisheries suggest that the average weight of mature Brights returning to the Columbia River has decreased approximately 2.7 kg since the 1910s, an average rate of about 0.1 lb·yr⁻¹ (45 g·yr⁻¹ ). Most of the potential biases in these data tend to make this estimate conservative. Insufficient data were available to describe changes in average age at maturity. There are many potential causes for the decline in average size of mature Brights, including factors that affect very early life stages. Other researchers have determined that size at maturity appears to be highly influenced by inheritance, gender, and growth rate. I describe how maternal size can influence -- through time of spawning, choice of spawning site, and egg size -- the viability of the young, which carry the dam's genes for size. The size-related ability to produce viable offspring may have been changed by modifications in the environment. Very little is known about how changes in the natural environment for spawning, incubation, and rearing may have contributed to a decline in average size at maturity. Artificial propagation and rearing, such as at Priest Rapids Hatchery, seems to produce adult Brights that are smaller, younger, and more likely to be male than their natural counterparts. The net result is that the average hatchery fish may have only about 0.80 of the reproductive potential of the average natural fish. Changes in growth conditions in the ocean probably did not contribute to the change in size, although the ocean fisheries of Southeast Alaska and British Columbia appear to select, in the genetic sense, against large size and old age in Brights. Since 1978, in-river commercial fisheries have caught larger Brights and a higher proportion of females than are found in the escapement of the Priest Rapids Hatchery component of the stock, but the fisheries impact the two sexes differently by taking the larger males and the smaller females. The effect on the natural component may differ because of their apparently larger average size. I found no evidence that larger fish or more females were caught when 8-in. minimum restrictions were in effect on gillnet mesh size relative to periods when mesh size was not restricted. Impounding the mainstem during the last 50+ yr may have removed obstacles to migration (e.g., Celilo Falls) that selected for large size in Brights, but that hypothesis could not be tested. The perserverance of larger and older phenotypes in the Bright stock suggests that countervailing selection -- perhaps during spawning, incubation, and/or early rearing -- may have resisted the effects of a century of size- and age-selective fisheries. That resistance, however, may reduce the productivity of the stock. Declines in average size and age at maturity can have undesireable consequences. Lower average size means less biomass landed and lower commercial value. Lower average fecundity and a diminished ability to reproduce in some environments are also expected. Loss of size and age classes may reduce the ability of the stock to adapt to environmental variations. These results are relevant to several management practices. A holistic approach to fishery management issues is necessary to avoid erroneous conclusions based on narrow perspectives. Measuring reproductive potential of the catch and escapement would be superior to the conventional practice of simply counting numbers of fish. Many aspects of artificial propagation can be improved, including broodstock aquisition, mating regimes, and rearing practices. Stock abundance is a major factor in determining the effect of many management practices on the stock. In general, fisheries managers must be mindful that they manage very complex natural systems. / Graduation date: 1992