The effects of 10 days prison solitary confinement on inmates' personal constructs and adrenocortical activity were examined. Eight Ss maintained their regular institutional routine. Eight Ss were placed for 10 days in solitary confinement. Personal construct (Repertory Grid Technique) rankings increased in stability for confined Ss as compared with non-confined Ss. This effect was more pronounced for "good" than for "bad" constructs for inmates independently rated as "simple" concept types. Adrenocortical function, as measured by plasma cortisol levels, indicated that solitary confinement was not more stressful than normal institutional life.Confinement in prison has been condemned on several counts (e.g., Mitford, 1973). In particular, the severest form of incarceration, solitary confinement, is looked upon as a cruel treatment since it is not only physically punishing but is often supposed to affect the inmates' selfidentity. However, while the solitary confinement of inmates does in some respects resemble sensory deprivation (Gendreau, Freedman, Wilde, & Scott, 1972), the literature (Zubck, 1969) is equivocal as to how "punishing" sensory deprivation is. The experiential set of prison inmates towards