The results of studies on association between the C677T polymorphism of the 5,10-methylene-tetrahydrofolate reductase (MTHFR) gene and osteonecrosis of the femoral head (ONFH) are controversial. To derive a more precise estimation of the relationship between the MTHFR C677T polymorphism and ONFH, a meta-analysis was performed. Eight studies on MTHFR C677T association with ONFH were searched up to April 2011, and the genotype frequencies in control group were consistent with Hardy-Weinberg equilibrium. The effect summary odds ratio (OR) and 95% confidence intervals were obtained. Publication bias was tested by funnel plot, Egger's regression test, and heterogeneity was assessed. Eight studies containing 778 cases and 1,162 controls were included. Heterogeneity was observed (χ(2) = 18.58, P = 0.01). Under the random effects model, the common OR was 1.38 (95% CI: 0.92-2.08; P = 0.12). In the subgroup meta-analysis, there was an association between MTHFR C677T polymorphism and ONFH in non-Asian population for CT + TT vs. CC (OR = 1.72; 95% CI: 1.21-2.43; P = 0.002; I(2) = 37.9%, P = 0.17), but not for Asian population (OR = 0.88; 95% CI: 0.66-1.66; P = 0.35; I(2) = 45.4%, P = 0.16). There was heterogeneity between studies and no clear evidence of an association on a worldwide population. When stratifying for the race, this meta-analysis did not provide an evidence of confirming association between MTHFR C677T polymorphism and ONFH. The large sample and well-designed study based on different ethnic groups should be considered in future associated studies to clarify the association of MTHFR C677T polymorphism with ONFH susceptibility.