During the last twenty years statistical methodology has become of central importance in research studies in medicine and also in day-to-day clinical practice. The medical literature is now liberally punctuated not only with relatively routine statistical terms such as p-value, t-test, confidence interval, and correlation, but also with more esoteric items such as hazard function, multilevel model, generalized estimating equations and crossover design. Consequently researchers in medicine and clinicians who are not primarily statisticians need to have a source that provides readable accounts of these terms so that they can understand at least the essence of the statistical aspects of both the design and analysis of a reported investigation. The Encyclopedic Companion to Medical Statistics is that source, containing readable accounts of over 500 statistical topics central to current medical research, with each entry being written by an expert in the field. Examples and graphical material supplement the written material in many entries, and extensive cross-referencing sign posts the reader to other entries that are likely to be relevant.