This paper reviews the methodology of studies of recent life events and illness, as relevant to psychosomatic studies. Three critical issues are discussed: the method of data collection; elimination of events which are consequences of illness; quantification of stress. It is concluded that the most important aspect is collection of data by careful interview technique covering defined events and time periods. Self-report event questionnaires, although heuristically valuable in the past, must now be regarded as inadequate. Exclusion of events which may be consequences of illness is also important in psychosomatic studies. Different methods of quantification produce refinements but are less crucial and alternative approaches may give similar results.