We study the fluctuations and correlations in a stochastic age-structured population model. Our model, which is closely related to certain "bit-string" models of evolution, incorporates survival probabilities that are density dependent, and also allows for varying reproductive probabilities as a function of age. We first solve for the simple steady-state of the deterministic version of our model, where all fluctuations and correlations are neglected. We then develop analytic techniques to calculate stochastic Gaussian corrections around this deterministic solution. This allows for a systematic, perturbative calculation of the population fluctuations and correlations. Away from the bifurcation point of the deterministic model we find good agreement with Monte–Carlo simulations.