The accuracy of available livestock data for the period 1960-85 have been assessed by examining the rationality of various assumptions under which such data were generated. The review shows that the aggregate data on livestock population and products have been generated under a set of inconsistent and often unrealistic assumptions with respect to calving, mortality, slaughter rates, import and consumption of livestock products. It Is suggested that available micro-level research and survey results should be used to formulate assumptions for generating macro data. Potential new research areas are also suggested.