In many medical schools biometrics has traditionally been taught with an abstract, mathematical emphasis. Consequently, the medical student usually does not relate the course content to his future practice as a physician, learns only to a limited degree, and develops a distaste for research methods and statistics. This unfavourable early exposure often produces a physician with little or no skill in assessing the merits of his professional literature in terms of research and statistical credibility. I therefore designed a course which emphasized research and statistical concepts, avoided mathematical explanations, stressed the relationship of research and statistics to medical practice, and used effective teaching principles. The course was first offered in the autumn of 1982 to second-year medical students at the Wright State University School of Medicine.