Harvest data constitute an important source of information for the deer manager, but interpretation can be complicated by reporting bias, hunter selectivity, differential vulnerability, and aging errors. Daily harvest records for >165,000 white-tailed deer (Odocoileus virginianus) taken during 1979-85 Illinois firearm seasons were examined for evidence of bias. Antlered males were apparently killed at a higher rate than other classes of deer; consequently, they became relatively less numerous in the herd and in the harvest as the season advanced. Yearling males, in particular, were highly vulnerable early in the season. Declining availability of antlered males and reduced hunter selectivity shifted pressure toward fawns and females as the season progressed. Fawns were apparently underrepresented in the total harvest but not the antlerless harvest. Male fawns were more vulnerable than their female cohorts. Biases involving various sex-age classes can affect calculation of certain population parameters from harvest data. The tendency toward increased representation of females with increased harvest intensity has implications for balancing recreational and population goals.