The need to defluoridate and fluoridate the water supplies in areas with drinking water naturally containing above–optimal (≥2.5 mg/l) and suboptimal (≤0.3 mg/l) fluoride concentration and caries and fluorosis prevalence of 12–year–old schoolchildren were assessed in Italy. In the low–fluoride area, 48.4% children were caries–free (DMFT = 0) and the DMFT and DMFS were 1.5 and 2.6; in the high–fluoride area, 46.8% had a DMFT = 0 and the values of the indices were 1.4 and 1.6, respectively. Multiple logistic regression analysis showed a significant association in the caries–free status according to parents' employment status (OR = 1.2, 95% CI = 1.1–1.3) and children's sweets consumption, since children who consumed sweets at least once a day had an adjusted odds ratio of 1.8 (95% CI = 1.4–2.3) compared to those with a lower consumption. Multiple linear regression analysis showed that DMFT and DMFS were significantly higher in children with a lower socioeconomic status and in those who consumed sweets at least once a day, with the DMFS significantly associated also with the area of residence. DT and FT scores were higher in the high– and low–fluoride areas, respectively. No evidence of fluorosis was reported in 94.5 and 55.3% of children in the low– and high–fluoride areas, respectively. The Community Fluorosis Index (CFI) for all permanent teeth was significantly higher in the high–fluoride area, 0.8, than the value, 0.1, found in the low–fluoride community. Our results substantiate the difficulties in defining universal guidelines for the fluoridation or defluoridation of drinking water and the need for an epidemiological approach to the decision as to fluoridate and defluoridate the water supply.