The fishery for yellow perch Perca flavescens in Green Bay, Lake Michigan, is currently operating under a rehabilitation plan based on a commercial harvest quota. We developed a bioeconomic computer model that included links between population density and growth, recruitment, and fishing effort for this fishery. Random variability was included in the stock-recruitment relation and in a simulated population assessment. We used the model in an adaptive management framework to evaluate the effects of the rehabilitation plan on both commercial and sport fisheries and to search for ways to improve the plan. Results indicate that the current quota policy is a member of a set of policies that would meet most management goals and increase total value of the fishery. Sensitivity analyses indicate that this conclusion is robust over a wide range of biological conditions. We predict that commercial fishers will lose money relative to the baseline condition, but they may receive other benefits from the elimination of the common-property nature of the fishery. The prospect exists for managing variability in harvest and stock size and for maximizing economic returns in the fishery, but more information is required, primarily on sportfishing effort dynamics and angler preferences. Stock-recruitment relations, density dependence of growth, and dynamics of sportfishing effort are the primary sources of uncertainty limiting the precision of our predictions. The current quota policy is about as good as other policies at reducing this uncertainty and appears, overall, to be one of the best choices for this fishery. The analytical techniques used in this study were primarily simple, heuristic approaches that could be easily transferred to other studies.