This paper reviews the potential of biometrics for the design of biological, medical and environmental research and, the interpretation of research results, and contrasts this potential with what has actually been achieved in improving standards of scientific research. It concludes that the successful development of biometrics will require us to look carefully at present and future practices in the planning and conduct of research, with particular emphasis on the responsibilities of senior research managers, on our communication of statistical theory, and on the use of computers.