In a study concluded in 1971, a consecutive series of patients admitted to Ward 1 of the Professorial Unit, Royal Edinburgh Hospital, had been assessed using psychological tests relevant to personality and to psychiatric illness (i.e. the Symptom-Sign Inventory, the Hostility and Direction of Hostility Questionnaire and the 16PF). The present study was an attempt to re-assess this cohort of patients three years after the key admission. The focus of interest was the relationship between personality and psychiatric illness. One of the most commonly emphasised distinctions between personality and illness is the relative persistence of the former and transience of the latter. By obtaining longer-term test-retest data, it was hoped to be able to identify satisfactory measures of personality and illness. Given such measures, it was intended to investigate whether personality, especially traits associated with psychopathy, had prognostic significance separable from that of illness. It was also intended to investigate whether personality could be shown to be relevant to illness aetiology. The main results were as follows: 1) The longer-term data on stability/change generally confirmed the findings of previous studies which had employed short test-retest intervals and so it was possible to identify measures of personality and illness. In addition, for some measures, previous contradictory evidence as to their personality/illness status was clarified. 2) Both personality and illness measures were predictive of illness outcome. However, personality measures related to psychopathy had associations with outcome separable from those of illness measures. The combination of marked degree of illness and deviance on traits related to psychopathy had associations with poor outcome. 3) None of the measures had relevance for the understanding of predisposition to become psychiatrically ill. The trait components of measures of intro-punitiveness and anxiety, and also some measures connected with extraversion, appeared to have relevance to predisposition to become more or less severely ill when 'breakdown' occurred.