Considerable uncertainty often exists in estimates of demographic parameters based on data collected from harvested furbearer species. We used molecular genetic techniques to estimate rates of error in 2 methods of sex determination of harvested bobcats (Lynx rufus): manual examination of the carcass (field sex) and laboratory-based maximum canine root area (MRA sex). Error rates were high for both sexing techniques, and were associated with age and an age—sex interaction for the field and MRA sexing methods, respectively. These findings do not support the use of the field methods for identifying sex of harvested bobcats. The MRA method may be effective for determining sex of older bobcats but is limited by considerable overlap between sexes in juveniles and yearlings. If critical demographic parameters are estimated from harvest data, efforts should be made to identify and reduce rates of error before data are used to assess population status.