The Cox's proportional hazards model indicated higher harvest levels significantly increased mortality of birth-year (BY) raccoons but not of animals ≥1 year old (ABY). Mortality of ABY males was significantly greater than that of ABY females. Hunters had a marginally significant (P = 0.10) effect on harvest mortality of BY raccoons but neither hunters nor trappers contributed significantly to harvest mortality rate of ABY raccoons. Activity and behavior patterns may contribute to greater harvest mortality of BY raccoons and ABY male raccoons;Monte Carlo simulations were done to investigate bias, efficiency and confidence interval coverage of the Kaplan-Meier (KM) and Heisey-Fuller (HF) survival estimators under a variety of conditions. Simulations explored robustness of HF to violation of a constant hazard rate in a time interval by allowing the hazard to decrease during an interval and by incorrectly identifying interval length. HF was not robust to a decreasing hazard unless survival was high during that interval. With low interval survival, HF had large negative bias and poor coverage. HF was partially robust to incorrect identification of interval length. Incorrectly identifying interval length or a decreasing hazard had little effect on KM. This indicates use of KM unless data clearly follow a piecewise exponential distribution with distinct intervals;Increasing harvest rate from a mean of 29% during 1983-85 to a mean of 48% in 1986-88 did not reduce population size. I found no density-dependent effects of increased harvest on pregnancy rates or litter size. Increased harvest seemed additive to natural mortality in BY raccoons but acted in a compensatory fashion among ABY animals.