This overview, based on a literature review and new data from the three cohorts (Whitehall Studies I and II, and the Vietnam Experience Study), has four objectives: (a) to outline the major determinants of height, so providing an indication as to what exposures this characteristic may capture; (b) to summarise, by reviewing reports from large scale studies, the relation between adult height and a range of disease outcomes--both somatic and psychiatric--with particular emphasis on coronary heart disease (CHD) and stroke; (c) to discuss why these relationships may exist, in particular, the role, if any, of socioeconomic position in explaining the apparent associations; and, finally (d) to outline future research directions in this field. The large majority of evidence for predictors of height, and its health consequences, comes from observational studies. While genetic predisposition is a major determinant of height, secular rises in childhood and adult stature across successive birth cohorts suggest that early life environment also has an important impact. Plausible non-genetic determinants of height include nutrition, illness, socioeconomic status, and psychosocial stress. Evidence for an association between height and a series of health endpoints is accumulating. Thus, shorter people appear to experience increased risk of CHD, and these associations appear to be independent of socioeconomic position and other potentially confounding variables. For stroke, and its sub-types, findings are less clear. In contrast to CHD, some cancers, such as carcinoma of the colorectum, prostate, breast (in women), central nervous system, skin, endometrium, thyroid and blood (haematopoietic) are more common in taller people. While height may be negatively related to the risk of completed suicide, conclusions about the links between stature and other health endpoints is problematic given the paucity of evidence, which should be addressed. Ultimately, for want of better data, investigators in this area have used height as a proxy for a range of pre-adult exposures. In future, research should aim to explore the predictive capacity of direct measures of diet, psychosocial stress, childhood chronic illness and so on, rather than focus on height or its components. The problem is that extended follow-up of child cohorts with such data are required, and studies which hold these data are not currently available, although several are either maturing to the point where they offer sufficient clinical outcomes to facilitate analyses or are in the advanced planning stage.