Limnological data collected at Lake Koocanusa were used to investigate the relationship of nutrient loadings, primary productivity, and trophic state of the reservoir during 1972-80. The reservoir, on the Kootenai River, was impounded by Libby Dam on March 21, 1972. Manipulation of the 7.16-cubic-kilometer reservoir for flood control, its primary function, created large fluctuations in reservoir volume and produced annual lake-filling times that ranged from 0.14 to 0.66 year. Loadings of nitrogen and phosphorus prior to and following impoundment of Lake Koocanusa were found to be large enough to predict eutrophic conditions. Beginning in 1976, total phosphorus loadings, but not total nitrogen loadings, were substantially reduced following improvements in waste-water treatment at a fertilizer plant located upstream from the reservoir. The closure of Libby Dam substantially reduced loadings of nitrogen and phosphorus downstream from Lake Koocanusa. On the average, the reservoir retained 63 percent of its influent loading of total phosphorus and 25 percent of its influent loading of total nitrogen. Daily areal and volumetric primary productivity varied widely in each year at four sampled limnological stations. During the 9 years studied, daily areal primary productivity, in milligrams of carbon fixed per square meter, ranged from 0.4 to 420.0; the mean of the 313 sampled days was 128.5. Annual areal primary productivity ranged from 23.2 to 38.5 grams of carbon fixed per square meter and thereby categorized Lake Koocanusa as oligotrophic. The relationship of annual areal primary productivity and 12 selected environmental variables was determined by multiple regression analysis. One of the models that was derived used two variables-annual euphotic zone depth and annual areal phosphorus loading-and accounted for 62.0 percent of the variation in annual areal primary productivity. The distribution of chlorophyll a within the water column indicated that, on the average, more than one-half of the phytoplankton in the reservoir was beneath the euphotic zone. These results support the hypothesis that the reservoir's weak thermal structure had allowed circulation of phytoplankton out of the euphotic zone. The trophic state of Lake Koocanusa was categorized as eutrophic when based on the relationship of the nutrient loadings and the reservoir's ratio of mean depth to hydraulic-residence time. This result conflicted with the oligotrophic ranking the reservoir received based on its areal primary productivity. The discrepancy in trophic state was attributed mainly to the failure of nutrient loading models to adequately account for physical processes within reservoirs. Part of the nutrient loading that entered Lake Koocanusa was unavailable to phytoplankton because the nutrients were carried beneath the euphotic zone by large volumes of interflow and underflow. Another part of the nutrient loading was adsorbed to suspended sediment and removed from the water column. Thus, phytoplankton primary productivity was controlled not only by nutrients, but also by other limno logical processes.