Around the world, a great deal of effort is expended each year to estimate the sizes of wild animal populations. Unfortunately, population size has proven to be one of the most intractable parameters to estimate. The capture–recapture estimation models most commonly used (of the Jolly–Seber type) are complicated and require numerous, sometimes questionable, assumptions. The derived estimates usually have large variances and lack consistency over time. In capture–recapture studies of long-lived animals, the ages of captured animals can often be determined with great accuracy and relative ease. We show how to incorporate age information into size estimates for open populations, where the size changes through births, deaths, immigration, and emigration. The proposed method allows more precise estimates of population size than the usual models, and it can provide these estimates from two sample occasions rather than the three usually required. Moreover, this method does not require specialized programs for capture–recapture data; researchers can derive their estimates using the logistic regression module in any standard statistical package.