Objective: Recent studies confirmed that resinmodified glass ionomers (RMGIs) set on the basis of two competing mechanisms, an acidbase reaction and a light-activated resin polymerization. This study evaluated the effect of the setting mechanism on bond strength by measuring the shear bond strength of three RMGIs to dentin with and without light activation. Methods: Sixty human molars were ground to midcoronal dentin and randomly divided into six even groups: 1) Ketac Nano (KN), 2) KN without light cure (woLC), 3) Fuji Filling LC (FF), 4) FF woLC, 5) Fuji II LC (FII), and 6) FII woLC. The dentin surfaces of the specimens were conditioned/primed according to the manufacturers' instructions. A 1.54-mm diameter plastic tube was filled with RMGI material and affixed to the dentin surface. Groups 1, 3, and 5 were light cured for 20 seconds, and groups 2, 4, and 6 were immediately placed in a damp dark box with no light curing at 378C for 24 hours. Shear bond strength testing was performed in an Instron device at 1 mm/min. Data were analyzed with a one-way analysis of variance (ANOVA) and Tukey/Kramer test (α=0.05). Results: Mean ± standard deviation shear bond strength values (MPa) are: 7.1 ± 4.2 (KN), 11.7 ± 3.9 (FF), 10.2 ± 3.2 (FF woLC), 12.5 ± 5.1 (FII), and 0.3 ± 0.4 (FII woLC). Two KN, all KN woLC, and seven FII woLC specimens debonded before testing. Tukey/Kramer analysis revealed no significant differences in bond strength between the three light-cured RMGIs. KN and FII showed significantly lower bond strength without light cure, but no significant difference was observed between FF and FF woLC. Conclusions: The results of this study strongly suggest that light activation is necessary to obtain optimal bond strength between RMGI and dentin. FF may contain components that chemically activate resin polymerization. Clinically, KN and FII need to be light cured after placement of these RMGIs.