OBJECTIVE To assess the one-year outcome of glass-ionomer cement (GIC) restorations with partial carious dentin removal in primary molars using digital subtraction radiographs. METHODS Children ages 6-8 years were recruited. Forty-nine primary molars with deep carious lesions were studied. The carious dentin was removed at the dentoenamel junction (DEJ) and restored with GIC. Digital radiographs were taken immediately after restoration, and at 6 and 12 months after restoration. The sets of digital radiographs were subtracted and analysed using Image-Pro Plus software. RESULTS At the 12-month follow-up, 38 of the original 49 primary molars were analysed. No postoperative pain or signs of pulp pathology were reported. 76% and 87% of molars showed an increase in density of the studied areas compared to control areas at 6- and 12- month follow-ups, respectively. The density of digital subtraction radiographs increased to 5.3 and 8.2 at 6 and 12 months from baseline, respectively. The mean density of the areas under restoration at 6 months and 12 months was statistically significantly higher than at baseline (paired t-test; p < 0.05). CONCLUSIONS Using digital subtraction radiographs, GIC restorations with partial carious dentin removal in primary molars showed a high potential for dentin remineralisation after 1 year.