Applicants for admission to St Mary's Hospital Medical School in 1986 were short-listed for interview by one of four assessors, who each made their assessments on a nine-item pro forma. One short-lister had also been studied in detail during 1981. Short-listers used the full range of possible judgements, in approximately the proportions requested. Only minor differences were found between them in the mean and range of their judgements, suggesting that similarity of standards can be maintained while using a number of separate short-listers. A confirmatory factor analysis of individual short-listers' judgements showed that all were extracting three separate factors, named 'Academic ability', 'Interests' and 'Contribution to community', although the less experienced short-listers differentiated these items less well than the more experienced. The short-lister assessed in 1981 and 1986 had retained an almost identical factor structure over the 5-year period.