This is a follow-up report on a research project dealing with the selection of psychiatrists for training. The original project was to have each candidate given three psychiatric interviews and a battery of tests interpreted by a psychologist. The four assessors made independent predictions of the applicant's suitability. A follow-up was then made in order to correlate the predictions and the findings. The results seem to indicate that using three interviews was wise since with just one interviewer his judgments are likely to be only slightly better than chance, whereas by pooling the findings of the