The general aim of this thesis was to evaluate whether a change in the threshold for surgical intervention in the caries process can be consistent with a stricter attitude towards the use of radiographic caries diagnosis. Bitewing radiographs of 3 groups of patients were retrospectively studied. Two groups comprised 229 patients, 18 years old at the end of the study in 1984. 102 had lived in an area with 1.2 ppm water-fluoride content (F84-group) and 127 in an area with 0.02 ppm fluoride content (O84-group). In the 3rd group 285 patients, 19 years old at the end of the study in 1993, lived in an area with 1.2 ppm water-fluoride content (F93-group). The latter patients were managed according to a restrictive attitude to surgical intervention and radiographic diagnosis of caries. The prevalence of patients and the frequency of tooth surfaces with caries was significantly lower in the fluoride groups than in the non-fluoride group. The correlation between patients' accumulated number of posterior proximal lesions and fillings at the last examination and the mean interval between their bitewing examinations was weak in all groups. The mean interval between examinations was significantly longer in the F93-group than in the other 2 groups. By applying an algorithm for individualisation of examination intervals these could be prolonged depending on the accepted risk for the development of inner dentin lesions. Future development of proximal dentin lesions was rather well predicted by means of past caries experience as demonstrated by ROC-analysis. The proportion of inner dentin lesions that were operatively treated was significantly higher in patients from the F93-group than in those from the other 2 groups in which a less strict attitude towards operative treatment was used. In the F93-group the average survival time of enamel and outer dentin lesions was 8.0 and 3.4 years, respectively, when right censored data were taken into account. In populations with low caries prevalence a strict threshold for operative treatment of proximal lesions can be consistent with prolonged intervals between radiographic examinations without an increased risk of pulp involvement. A skewed distribution of the patients with regard to caries experience was found in all 3 groups. Thus, even in low prevalence populations the intervals between radiographic examinations must be individually determined and coupled with individualised preventive care.