OBJECTIVE This study evaluated in situ the potential of a glass ionomer and self-adhesive resin cements to inhibit enamel and dentin demineralization around indirect restorations exposed to cariogenic challenge. The cumulative fluoride release (CFR) of materials was measured in water and acid. METHODS Seventy blocks cut from human molars received two indirect composite restorations (one in enamel and another in dentin) luted with Ketac Cem EasyMix (GIC, positive control), SeT (SeT), Maxcem Elite (Max), Smart Cem2 (Smart), and RelyX Unicem 2 (Unicem2). Fourteen volunteers wore palatal appliances containing five blocks exposed to a cariogenic challenge (20% sucrose solution, eight times per day, seven days). Knoop microhardness (KH) at two distances from the margins and three depths from the outer surface determined enamel and dentin demineralization. Disc-shape specimens of materials were immersed in daily-replaced deionized water or lactic acid solutions. KH and CFR data were analyzed by analysis of variance, Games-Howell test, and Tukey test (α=0.05). RESULTS The overall KH ranking was GIC > SeT > Max > Smart = Unicem2 in both enamel and dentin (">" means p<0.05). SeT was the only resin cement that resulted in enamel and dentin KH comparable to that of GIC at most distances and depths. In water, CFR rank of materials was GIC > SeT = Max > Smart = Unicem2. In acid, the rank was similar, except that Set was significantly superior to Max. CONCLUSIONS SeT inhibited demineralization in enamel and dentin quite comparably to GIC. All resin cements released lower cumulative amounts of fluoride than the glass ionomer cement.