Pulpal reaction to restorative materials depends on marginal microleakage, which is a dynamic phenomenon that allows bacteria and fluids to traverse across the tooth-restoration interface. Glass-ionomer cement (GIC) exhibits low microleakage due to direct bonding to tooth structures. Hydroxyapatite (HAP) based on the similarity with tooth structure may decrease the microleakage. The aim of this in vitro study was to evaluate marginal microleakage of a mixture of conventional and resin-modified glass-ionomer (RMGI) with micro- and nano-HAP. In this in vitro study, 30 non-carious extracted human third molar teeth were used. Standard Cl V cavities were prepared on the buccal and lingual surfaces. The cavities were restored in six experimental groups as follows: group 1, conventional glass-ionomer cement (CGIC); group 2, CGIC with micro-HAP; group 3, CGIC with nano-HAP; group 4, RMGI; group 5, RMGI with micro-HAP; group 6, RMGI with nano-HAP. The restorations were finished and polished. The teeth were coated with nail polish, sealed with sticky wax, thermocycled and placed in a solution of 2% basic fuchsine for 24 hours. The teeth were sectioned and microleakage was measured. Kruskal-Wallis, Man-Whitney and Wilcoxon tests were used for data analysis (P<0.05). The data analysis revealed significantly lower microleakage in groups 5 and 6 at both occlusal and gingival margins. Also in these two groups the gingival microleakage was significantly lower than occlusal margin (P=0.009 and P=0.001 respectively), but in groups 1(CGIC) and 3(CGIC+ nano-HAP) and 4(RMGI) the microleakage of occlusal margin were significantly lower than that of gingival margin (P=0.001, P=0.007 and P=0.001 respectively). Mixing RMGI with nano-HAP and micro-HAP resulted in lower microleakage.