Penguin populations are potentially sensitive indicators of change in the Antarctic and sub-Antarctic and are used extensively to monitor changes in the marine ecosystem. Aerial photographic surveys may provide the most robust method for estimating breeding population size, particularly for large colonies, for colonies in areas of complex terrain, or for colonies where ground access is difficult. Obtaining population estimates from aerial photographs is laborious and time consuming, and is usually carried out manually by counting individual birds on highly magnified prints. Here we present the results of a computer based image analysis of digitally scanned colour aerial photographs of macaroni penguin colonies on Bird Island, South Georgia. We compare the results with conventional ground counts for the study colonies, highlighting various assumptions that contribute to differences in the population estimates. The software analysis method provides population estimates that are considered to be more reliable than the associated ground counts. The technique potentially provides a reliable method for carrying out large-scale wildlife management surveys.