Game counts provide essential information for the management of wildlife populations. On Ongava Game Reserve, two methods are used annually to count animals: a helicopter aerial count and a 72 hour waterhole count. During the extended waterhole count, observers remain in place for the duration of the count, and are required to record details of all animals seen. In this study we assessed whether camera traps could assist in the count process, firstly by easing the vigilance burden of observers in the nocturnal periods, and secondly by improving the accuracy of the overall count. We found that camera traps not only can substitute for human observers for nocturnal counts, they record more events and hence count more animals. We also found that traps record cryptic and small species that are often missed by observers. Ongava now uses camera traps as the primary counting method from midnight to dawn. In contrast, we found that camera traps cannot substitute for observers for diurnal observations when animal group sizes of more than ten animals are present. Even with extended trap arrays, the field of view and complexity of individual movements in large groups compromises the ability of offline counters to track animals, and they underestimate group sizes. However, there remain a number of advantages of running a camera census in parallel with observers - sightings can be checked and verified, and traps have a consistent mode of operation, minimising inter-observer bias. Images of marked animals can be reviewed offline for further analysis. We believe that the additional information recorded by camera traps provides important population data for both prey and predator species, allowing the development of an integrated ecosystem management strategy. Keywords: camera trap, game count, Namibia, nocturnal animals, waterhole.