A series of studies have explored the role of cytosolic serine hydroxymethyltransferase (SHMT1) C1420T polymorphism in cancer risk, but their results were conflicting rather than conclusive. To derive a more precise estimation of the association between C1420T and cancer risk, the present meta-analysis of 28 available studies with 15,121 cases and 18,023 controls was conducted. The results revealed that there was no significant association between the polymorphism and cancer risk overall. In stratified analysis by cancer type (breast cancer, gastrointestinal cancer, leukemia, lymphoma, and others), the results showed that 1420T allele was associated with decreased risk in leukemia (CT vs. CC: OR= 0.825, 95% CI =0.704-0.966; and CT+TT vs. CC: OR= 0.838, 95% CI = 0.722-0.973), but the same results were not present for other cancer types. When subgroup analysis was performed by source of control (population-based [PB] and hospital-based [HB]), a borderline inverse association was observed for the HB subgroup (CT vs. CC: OR= 0.917, 95% CI = 0.857-0.982) but not for the PB subgroup. Stratifying by geographic area (America, Asia and Europe), significant inverse association was only found in Asia subgroup (CT vs. CC: OR= 0.674, 95% CI = 0.522-0.870). In summary, the findings suggest that SHMT1 C1420T polymorphism is not associated with overall cancer development, but might decrease cancer susceptibility of Asians as well as reduce leukemia risk. Large well-designed epidemiological studies will be necessary to validate the risk identified in the current meta-analysis.