Apart from being used as diagnostic instrument in clinical settings the Present State Examination is also employed in open populations to assess the prevalence of psychiatric symptoms. In a pilot study we examined two aspects of the reliability of the PSE: its interraterreliability and its reliability in distinct populations. With regard to the first aspect independent judges showed a fair amount of agreement. The reliability proved not to be the same in the four patient categories. With the patients of general practitioners and with those from a clinic for neurotics the interview reliability was better than with patients from a general mental hospital. With non-patients too the reliability was acceptable. Our results tend to confirm internationally known research findings. The PSE proves to be adequate for research in a general population.