The food habits, survival rates, and abundance of four predatory fish species (striped bass, walleye, largemouth bass, smallmouth bass) were determined and combined with information on energetics and growth to quantitatively assess predator-prey interactions in Smith Mountain Lake, a large hydroelectric reservoir in south-central Virginia. A total of 1871 stomachs of the four species was examined from fish collections between April 1983 and December 1984. Food habits of each predatory species were assessed and compared by season using Schoener's (1970) diet overlap index. Significant diet overlap occurred between the pelagic predators (striped bass and walleye) and between the littoral black basses (largemouth bass and smallmouth bass) in all seasons and locations where sufficient stomach samples were collected. Survival and abundance estimates for the four predators were derived from several data sets. First-year survival of striped bass fingerlings averaged 20.8% for an 11-year period (1973 to 1983) and ranged from 3.9% to 54.3%. Average survival of age-0 largemouth bass and smallmouth bass was 53% and 45%, respectively, for the period 1979 to 1983. Species-specific survival rates for predators age 2 and older ranged from 29% to 59%. Of the predators age 1 and older, largemouth bass were the most numerous at an estimated density of 61.2/ha, followed by striped bass at 16.3/ha. Age 1 and older smallmouth bass and walleye were less abundant, at densities of 8.2 and 2.0/ha, respectively. A bioenergetics model was developed for striped bass and, with models developed for walleye and largemouth bass, used to estimate annual consumption of alewives and gizzard shad by the four piscivorous species. Annually, striped bass consumed the greatest amount of adult and young-of-year alewives, whereas largemouth bass ate nearly 70% of the total shad biomass consumed by all predators. A comparison of predator demand to clupeid supply substantiated diet overlap data; namely, that native and introduced predators were compatible in Smith Mountain Lake. The striped bass bioenergetics model was used to make inferences about potential stocking strategies for striped bass based on availability of age-1 and older alewives. A stocking rate of 20-24 fingerlings/ha appears to maximize production of striped bass and minimize potential competitive interactions between striped bass and native black bass in Smith Mountain Lake. / Ph. D.