Although wildlife management agencies commonly employ sex-selective harvests to regulate white-tailed deer (Odocoileus virginianus) populations, few studies have documented the effectiveness of these harvests. Using data from 1980 to 1997 for the Algonquin Highlands region of Ontario, Canada, we assessed (1) the ability of wildlife managers to control the size of the antlerless harvest using sex-selective permits, and (2) the ability of antlerless harvest to control changes in deer density. Antlerless harvest was related only to the number of permits issued when <40% of hunters had antlerless permits; above this threshold, kill was related only to hunter numbers, not the number of antlerless permits. Factors such as deer encounter rates and hunter selectivity or behavior also may influence the size of the kill. Historically, antlerless kill showed little detectable effect on deer population density, which appears to be regulated primarily by density-dependent factors. This implies that antlerless kill historically occurred at levels too low to depress populations, or that existing data are simply too noisy to allow detection of a kill effect. Either way, the current harvest management system appears to have little ability to regulate deer populations in Ontario. Declining hunter participation and/or increasing deer populations will only decrease the effectiveness of the current sport harvest for management, and wildlife managers may need to look to other means of managing the population.