Forecasting changes in size and distributions of populations is an essential component of conservation assessments. Such forecasts are only useful for species conservation and management when they are based on robust estimators of fecundity, survival, and density dependence. While apparent survival estimation is the main focus of mark–recapture modeling, fecundity and density dependence are rarely the subject of these models. Here, we present a Bayesian hierarchical framework that can estimate fecundity and density dependence along with age-based survival using only robust-design capture–recapture data. We refer to this framework as RD-pop. We used simulated capture histories to demonstrate that RD-pop can estimate vital rates and their density dependence with little bias. We applied RD-pop to capture history data from Brown Creeper (Certhia americana) and showed that estimates of fecundity are consistent with the breeding biology of this species. Finally, we illustrate that density dependence, even when estimated with uncertainty in the RD-pop framework, regularizes population dynamics and reduces the frequent population extinctions and explosions observed under density-independent models. RD-pop is a useful addition to the current mark–recapture modeling toolbox especially when the goal is to build population models that can make medium- and long-term projections. It can be applied to any population for which long-term robust-design mark–recapture data are available, and with slight modifications (incorporation of weather and climate effects on vital rates) has the potential to facilitate demographic projections under climate change.