The primary value of bioassays is that they alone directly assess biological activity of bioactive substances and products. Appropriately designed bioassays reflect the fundamental aspects of the biological activity of a bioactive molecule, including ligand-receptor binding, signal transduction processes (often poorly understood) and the final observed biological effects. Biological assays therefore complement physicochemical and biochemical procedures which normally only assess precise molecular structural features of complex molecules produced by biotechnology. Bioassays provide valuable information concerning the potency of biological products. This is essential for evaluating batch-to-batch consistency and stability. Bioassay data are crucial at all stages in the development of biological products, from early research work to final quality control of finished products. However, the type and design of bioassays may differ according to the information required and its intended use. The assays may or may not directly relate to the clinical use of the product. Bioassays can be difficult to perform and be time consuming, although this often reflects bad assay choice and/or design. Correct analysis of the assay results is essential if valid data are to be obtained. Standardisation, using correctly calibrated primary and secondary standards, is essential. In vivo bioassays are normally more unreliable than in vitro procedures, but in some cases the latter are either not available or do not address important biological characteristics of a product. Bioassays must be validated for their intended purpose and for the types of samples to be measured. Appropriate statistical analysis should be used to derive the significance and specifications of results. This needs to address both variability in samples and assay performance. Specifications (limits) for product acceptability need to be derived from real data using several batches of product.