Largemouth bass Micropterus salmoides populations declined in Kentucky and Barkley lakes following a severe drought during Dec 1984–Oct 1988 and below normal rainfall in three of four years during 1989–1992. Regression models developed for both reservoirs indicated that recruitment of largemouth bass to age 1 was positively related to the electrofishing catch per unit effort (CPUE) of large age-0 bass and negatively related to annual rainfall amounts and the CPUE of all sizes of age-0 largemouth bass. These variables explained 73% of the variation in CPUE of age-1 bass at Barkley Lake and 60% at Kentucky Lake. We also developed regression models to estimate the recruitment of largemouth bass to the fishery at age 5 and older. Recruitment to the fishery at Kentucky Lake was positively related to the CPUE of age-1 bass and negatively related to reservoir discharge. This model explained 52% of the variation in recruitment to age 5 and older largemouth bass. Recruitment to age 5 and older fish at Barkley Lake was positively related to the CPUE of age-1 bass and rainfall and negatively related to reservoir discharge. The model developed at Barkley Lake explained 82% of the variation in recruitment of largemouth bass to the fishery. The models predicted that the largemouth bass fisheries at both reservoirs would decline through 2001 under the current angling regulations, which include a 356-mm minimum length limit and one fish less than 356-mm total length allowed in the angler's creel. We used these models, in consultation with the public, to change length limits to a 381-mm minimum length limit in an attempt to moderate the expected declines in older fish.