Maximizing sampling efficiency and reducing sampling costs are desirable goals for fisheries management agencies. Expensive and labor-intensive methods (such as mark-recapture) are commonly used to estimate the population abundance of walleye Sander vitreus, but more efficient methods may be available. We compared recapture rates from surveys and harvests to evaluate the efficiency of currently used recapture gears and the potential for using gears that require less effort. To evaluate the usefulness of walleye harvest as mark-recapture samples, we used errors-in-variables models to determine whether recapture rates differed between fyke-netting and spearing, electrofishing and spearing, and electrofishing and angling. We found no significant differences between fyke-netting and adult walleye electrofishing recapture rates or between spearing and adult walleye electrofishing recapture rates. In contrast, we found that recapture rates from angling and electrofishing differed significantly in lakes with and without minimum length limits. We concluded that the lack of significant differences between the slopes of some harvest and survey recapture rates may allow the use of harvest recapture rates to estimate walleye abundance, but the biases associated with each gear should be considered. We also concluded that more attention should be given to understanding the biases of recapture gears. ?? Copyright by the American Fisheries Society 2005.