Summary.1Three groups of students applying for admission to a university education department for teacher training were interviewed. Interview ratings were made under a number of different headings by the interviewers.2These interview data, on analysis, indicated that three factors were operating: (a) a factor based on the observable features of the applicant; (b) a factor based on judgments of this intellectual maturity and his ability to express himself; (c) a factor based on a number of personality traits thought to be important in teaching.3Follow‐up data based on teaching marks indicated that the first two factors above were those which gave the best prediction of teaching ability.4The sizes of the follow‐up correlations suggested that the interview, provided that it was not expected to yield information beyond its powers, was as good a predictor of success at this level of education as most other measu