We used cohort analysis to reconstruct the female segment of a Columbian black-tailed deer (Odocoileus hemionus columbianus) population from 1979 to 2000 in the western Cascades of Washington, USA. We used reconstructed population estimates and age class representations to analyze relationships between population change and female density, forage availability, and weather influences. We applied stage structured and unstructured modeling approaches and used information-theoretic methods to select the best models. We used habitat covariates to develop predictive functions for fertility and survival parameters in structured models. The best structured and unstructured models were composed of combinations of factors including population density, forage availability, and winter weather. Structured and unstructured models could assist with management of black-tailed deer by providing the ability to predict deer population change given covariate values.