The basis for whole animal bioassays as practised today was laid down by Paul Ehrlich in 1894. He introduced the concepts of a stable standard preparation and of the unit of activity as the activity of a defined mass of that standard preparation in the assay performed. Such assays have often provided a way of quantifying newly discovered active principles of biological origin, so that they could be applied in clinical medicine. Whole animal bioassays can be applied not only for the quantitative analysis of a biological product (analytical assays), but also for the comparison of different products intended for the same clinical indication (comparative or research assays). As such they have been the model for controlled clinical trial. For some products many different types of bioassay have been developed. They may produce heterogeneous results when more than one active principle is involved, and these are present in standard preparation and in the unknown preparation in different relative concentrations. In addition the precision of different assay methods for the same substance may vary markedly. An important source of variation in whole animal bioassays is the influence of some environmental factors on the individual animals during the assay. Thus the number of rats per cage markedly influences the variation of the response of rats to serum gonadotrophin. Careful studies are required to detect the discriminatory environmental factors in a particular whole animal bioassay. Keeping these discriminatory factors constant at their optimal level may increase the precision of the assay markedly and thus reduce the number of animals required to attain the precision specified, e.g. in pharmacopoeial tests.