Since the mid-1980s Ian Goodyer and his colleagues have been engaged in trying to understand the dynamics of the relationship between acute and chronic life stresses and depression in adolescents (Goodyer et al. 1985, 1986). One of the great strengths of their work has been their attention to an array of potential risk factors ranging from the apparently exogenous (such as life events, e.g. Goodyer et al. 1988; Goodyer & Altham, 1991; Goodyer, 1999), through those of uncertain origin (e.g. peer and family relationships and problems, Goodyer et al. 1990), to those that can be regarded as being endogenous to the child (e.g. measures of temperament or endocrine function, Goodyer et al. 1993, 1996, 1998; Herbert et al. 1996). They have then typically attempted to determine how subsets of these factors interact in the generation of disorder (see Goodyer, 2002 for a speculative presentation of his current thinking about the relationships between social adversity and depression). A second characteristic of their research lies in its frequent attention to outcomes over time, as opposed to cross-sectional relationships (e.g. Goodyer et al. 1997a,b). Thirdly, in line with growing evidence that the relationships between environmental stress (and perhaps neuro-anatomical correlates) and depression change as repeated episodes occur (Lewinsohn et al. 1999; Daley et al. 2000), they have distinguished among first, recurrent, remitting and chronic depressive episodes. The group's latest work, presented in this issue (Goodyer et al. 2003), nicely illustrates these characteristics, and their extension to consideration of potential dynamic relations among endocrine parameters in relation to psychopathology.