Three hundred and twenty-five psychiatric patients were allocated to classes within the hierarchy of personal illness by means of the Delusions-Symptoms-States Inventory. They were then given the Personality Deviance Scales. The results showed that the classes ranked in the same hierarchical order as on the DSSI on extrapunitiveness and intropunitiveness, but not on dominance. Maladjustive personality deviance, as statistically defined, was 3 1/2 times as frequent among patients as among non-patients. Whereas symptom measures had previously been shown to change considerably after one month, personality measures did not. Longer follow-up periods are needed before it can be decided whether personality measures contribute substantially to prediction of type of illness or whether they are determined, at least in part, by the type of illness. In the latter event, personality measures might still prove useful in providing a more fundamental estimate of long-term clinical improvement than symptom measures alone if they were found to change more slowly or only with more intensive clincial effort.