To observe the chronic combined effects of sodium fluoride and sodium arsenite on the Runx2 and downstream related factors of bone metabolism in SD rats. SD rats were divided randomly into nine groups of 6 each by factorial experimental design (half female and half male) , and supplied with the different doses of fluoride, arsenite and fluoride plus arsenite containing in deionized water (untreated control containing 0 mg/kg fluoride and 0 mg/kg arsenite, and low-fluoride and high supplemented with 5 and 20 mg/kg fluoride, and low-arsenite and high supplemented with 2.5 and 10 mg/kg arsenite, and low-fluoride plus low-arsenite, and low-fluoride plus high-arsenite, and high-fluoride plus low-arsenite, and high-fluoride plus high-arsenite, respectively) . After 6 months exposure, the concentration of Runx2, matrix metallopeptidase 9 (MMP-9) ,Osterix, Receptor activator for nuclear factor-κ β ligand (RANKL) were detected by enzyme-linked immunosorbent assay method, respectively. There were no dental fluorosis found in the control group, low-arsenic group and high-arsenic group. There were significant differences in the constituent ratio of dental fluorosis among the rats from low-fluoride and high-fluoride (that is 5 rats out of 6 and 6 rats out of 6) compared with the control group (0 rat out of 6) (χ(2) = 8.57, 12.00, P < 0.05). The bone fluorine level increased with the increase of fluoride dose, the groups without fluoride supply (control group, low-arsenite and high-arsenite group's geometric mean (minimum-maximum) were 0.005 (0.003-0.009), 0.006 (0.003-0.021), 0.003 (0.002-0.100) mg/g, respectively), low-fluorine groups (low-fluoride group, low-fluoride plus low-arsenite, and low-fluoride plus high-arsenite group were 3.395 (2.416-5.871), 3.809 (1.471-7.799), 1.471 (1.473-6.732)mg/g, respectively) , the high-fluorine groups (high-fluoride, high-fluoride plus low-arsenite, and high-fluoride plus high-arsenite group were 70.086 (46.183-131.927), 69.925 (40.503-96.183), 40.503 (52.622-89.487) mg/g, respectively) and the differences between groups was significant (P < 0.05). The bone arsenic level increased with the increase of arsenite dose. The low-arsenic groups (low-arsenite group, low-arsenite plus low-fluoride, and low-arsenite plus high-fluoride group were 7.195 (5.060-9.860), 6.518 (2.960-12.130), 6.970 (3.400-9.730) µg/g, respectively), the high-arsenic groups (high-arsenite, high-arsenite plus low-fluoride, and high-fluoride plus high-arsenite group's geometric mean(minimum-maximum) were 8.823 (5.760-10.840), 9.470 (7.230-12.860), 8.321 (2.420-17.540) µg/g, respectively) were significantly higher than that in the groups without arsenic supply (control group, low-fluoride and high-fluoride group were 1.785 (0.300-3.750), 2.226 (1.410-3.980), 2.030 (1.040-3.850)µg/g, respectively) (P < 0.05). There was no significant difference of the bone arsenic concentration between low-arsenic and high arsenic group. There was significant positive correlation between fluoride concentration and Runx2, MMP-9, Osterix, RANKL level (the correlation coefficient was 0.647, 0.354, 0.582, 0.613 between fluorine gavage concentration and protein level, the correlation coefficient was 0.559,0.387, 0.487, 0.525 between bone fluorine concentration and protein level, respectively, P < 0.01). There was negative correlation between arsenite gavage concentration with Runx2 level (r = -0.527, P < 0.05) and was no correlation between arsenite gavage concentration with MMP-9, RANKL,Osterix level (P > 0.05). There was interaction between fluoride and arsenite to Runx2, MMP-9, RANKL,Osterix (F = 3.88, 15.66, 2.92, 6.42, respectively, P = 0.01, <0.01, 0.031, <0.01, respectively). The combined effects of fluoride and arsenic on the Runx2, MMP-9, RANKL, Osterix of bone metabolism showed antagonistic effects.