The object of the present study was to determine the relationship between dental attendance and the levels of untreated caries in a sample of 11-12 year old pupils attending first form of second level schools in the South Belfast area of Northern Ireland. A sample of 199 first form pupils attending second level schools in the area was obtained. The diagnosis of caries depended mainly on visual evidence following removal of debris and moisture. Blunted probes were used to confirm or reject doubtful lesions and no radiographs were taken. Details of parental employment status were obtained by a questionnaire to the parents. The data were entered on to an N.C.S.S. data analysis system for stepwise regression and analysis of variance. Overall there was a significant difference in the untreated caries status of those with different attendance patterns, the major difference being that a much higher proportion of those who had not attended within the previous 12 months had active caries. Almost half of the children had attended in the previous six months and only 16 (8 per cent) had not attended in the previous 12 months. Independent analysis indicated that parental employment status was significantly related to the level of untreated caries (p > or = 0.05). Independent variables entered in the stepwise regression were attendance, employment status and DMFT. The strongest predictors of untreated caries were previous caries experience and period since dental attendance, while employment status was a significant but weaker factor in the analysis. These findings may be relevant when planning the frequency of school dental screenings and managing the frequency of dental recalls and examinations.