The utility of consumer-satisfaction data in program-evaluation studies is inspected by means of direct and indirect patient evaluations of treatment effects and of personal functioning. These evaluations are collected from 656 ex-clients 3 years after discharge by means of a self-report questionnaire. The subjects were treated at either PTC's for poorly integrated patients or PTC's for the treatment of severe neurotic disorders. Their evaluations prove to be less positive than those collected from a `traditional' psychiatry population, although there is a sample bias effect as far as the less satisfied patients are concerned. Data obtained from the neurotic population render useful information on the evaluation of treatment effects. Self-attribution mechanisms may account for less positive evaluations. The validity of the evaluations obtained from the psychotic group may be questioned as these patients prove to perceive their functioning apart from readmissions.