Twelve duplicated bitewing radiographs of clinical quality were viewed under controlled conditions by 42 dentists in a teaching department. The dentists noted surfaces with initial caries or with caries in need of treatment. The radiographs were rediagnosed by 20% of the dentists to determine inter- and intra examiner reliability. A norm for the validity of lesion presence in enamel or dentin was applied. The variation in surfaces indicated for treatment is illustrated by: mean 40, s.d. 16, min. 13 and max. 89 for the 42. The mean proportion of true positive filling decisions in percentages was 67, s.d. 18, the mean proportion of false positive filling decisions was 22, s.d. 22. A large individual variation was found in this study. The dental teachers saw less caries present than the norm but some indicated more fillings needed than dentin caries present. It is suggested that care in the educational process should be given to developing appropriate caries treatment criteria. Factors conducive to preventively orientated treatment decisions in the clinical situation need further attention.