The Nuffield Unit of Tropical Animal Ecology was set up to study tropical animals in relation to their environment. The initial studies were aimed at establishing the population densities of large mammals and their interactions with the environment, especially vegetation, climate and soils, and with other animals. At an early stage of the research programme a series of study areas was selected, with a view to obtaining regular quantitative information on the numbers, densities, distribution and population structure of the larger animals, and to relate this information to data on the vegetation and other environmental factors. Preliminary observations had shown that the overall distribution of mammals in the Park was uneven and locally showed considerable heterogeneity in different vegetation types both in absolute numbers and relative to other species. Twelve easily accessible study areas were chosen which were considered to be generally representative of the varied grassland ecosystems of the Queen Elizabeth Park, as regards different animal associations, vegetation communities, standing water and varied distance from the lakes and the Kazinga Channel; an additional study area was sampled for 1 year only. Monthly counts of animals began in March 1963 and continued to Macrh 1967. The species involved were elephant (Loxodonta africana Blumenbach), hippopotamus (Hippopotamus amphibius Linn.), warthog (Phacochoerus aethiopicus Pallas), buffalo (Syncerus caffer Sparrman), Uganda kob (Adenota kob Neumann), waterbuck (Kobus defassa Ruppell), reedbuck (Redunca redunca Pallas), bushbuck (Tragelaphus scriptus Pallas) and topi (Damaliscus korrigum Ogilby). It was also intended to relate expected changes in the densities of grazing animals to changes in the number of hippopotamus induced by management cropping in some of the study areas. Management cropping involved shooting and utilization of the carcases and is referred to below as 'cropping'. Owing to our limited resources and the demands of other programmes it was not possible to determine primary productivity or standing crop of plants, and the vegetation survey was limited to mapping the distribution of plant communities in the study areas. The results from ten of the thirteen study areas, in which similar counting techniques could be applied, are presented. Scientific names follow Ellerman, Morrison-Scott & Hayman (1953),l Harker & Napper (1960) and Dale & Greenway (1961).