This study was carried out in order to determine the extent to which ions released from fluoride-containing dental restoratives migrated through the enamel and dentine of extracted teeth. A total of 40 permanent human 3rd molars were used. They were extracted for orthodontic reasons, and employed within 1 month of extraction. A cervical (Class V) cavity was prepared in each tooth, then filled with one of: a conventional glass-ionomer, a resin-modified glass-ionomer, a polyacid-modified composite resin ("compomer") or a fluoride-releasing resin composite. Ten samples were prepared per material. After 1 month, five specimens per material were prepared and examined under SEM/EDX. Concentrations of sodium, aluminium, strontium, fluorine, magnesium, silicon, phosphorus and calcium were determined within the tooth. After 18 months, the remaining five specimens for each material were prepared and studied in the same way. The greatest extent of ion migration into the tooth was found with the conventional glass-ionomer and least migration was found for the fluoride-releasing composite, which showed no evidence of fluoride migration at all. Levels of migrating ions were generally higher in the 18 month specimens than in the 1 month specimens, and also higher in the dentine than in the enamel. Ions released by restorative dental materials have been shown conclusively for the first time to be capable of migrating into the enamel and dentine surrounding the restoration. The conventional glass-ionomer showed the highest level of ion migration whereas the fluoridated composite resin showed little if any ion migration. This suggests that the conventional glass-ionomer has the greatest caries inhibiting effects of all the materials tested, and the fluoridated composite the least.