Glutathione S-transferases (GSTs) are postulated to be involved in the detoxification of potential carcinogens, and gene variation may alter susceptibility to lymphomas. Results from several previous epidemiologic studies have been inconsistent. Hence a meta-analysis was conducted to verify the role of GST genetic polymorphisms in lymphoma risk. Eleven trials involving 1626 patients and 2892 controls were analyzed. Pooled results showed that the GSTT1 null polymorphism might increase the risk of lymphoma (odds ratio [OR] 2.26, 95% confidence interval [CI] 1.20, 4.24; p = 0.01; random-effects model), whereas the impact of GSTM1 and GSTP1 Ile105Val polymorphisms was not significant. Subgroup analysis showed the GSTT1 null genotype to be a risk factor for non-Hodgkin lymphoma (NHL) (OR 2.75, 95% CI 1.17, 6.45; p = 0.02; random-effects model) but not for Hodgkin lymphoma (HL), and the effect remained evident in females (OR 1.43, 95% CI 1.04, 1.97; p = 0.03; I(2) = 41.0%, p for heterogeneity = 0.15; fixed-effects model). An effect of GSTM1 and GSTT1 double null genotype on lymphoma risk was also shown (OR 2.09, 95% CI 1.31, 3.33; p = 0.01; random-effects model). In conclusion, the GSTT1 null genotype appears to be associated with a modest increase in the risk of NHL, whereas the GSTM1 and GSTP1 Ile105Val polymorphisms are unrelated to lymphoma risk.