An array of models for sustainable harvesting has been proposed, and most are in some way dependent on a density-dependent response. I tested a subset of these models by manipulating three North American deer populations to reveal density dependence and determine population response to harvests. These were a white-tailed deer Odocoileus virginianus population on the George Reserve (GR) in southeastern Michigan, and black-tailed deer O. hemionus columbianus populations on Hopland Research and Extension Center (HREC) in Mendocino County, California, and on Fort Hunter Liggett (FHL) in Monterey County, California. The study areas represented a gradation in size, productivity, management control, and environmental stochasticity. The GR population showed a nearly linear density-dependent relationship of r on N. A mean maximum sustainable yield (MS Y) of 49 deer/year was obtained at an N of 56% of K carrying capacity (KCC). A fixed harvest with stochasticity, however, reduced the sustainable MSY to 43 deer/year. A second population growth experiment 50 years after the initial introduction showed an equivalent growth rate. Analysis of harvest data from FHL showed buck harvest to be positively related to size of previous female harvests. From these results, a ‘linked-sex harvest strategy’ (LSHS) was proposed in which female harvest is sequentially incremented so long as buck harvest continues to increase, up to a presumed ‘safe’ female to buck ratio. At HREC, bucks were harvested in public bucks-only seasons. Buck harvest was monitored for 6-year pre- and posttreatment periods without female removals, and for a 7-year treatment period during which 20 females/year were removed for three years and 30 females/year for four years. There was no significant difference in pre- and post-treatment period buck harvests, so they were combined as a ‘control’. There was a significant (25%) increase in buck harvest during the treatment period despite its coinciding with six consecutive years of drought. The combined-sex harvest was more than double that of the buck harvest alone during the control periods. The relevance of these studies to deer harvest management is discussed.