Many studies have examined the interaction between CYP1A1 MspI gene polymorphism and smoking for the risk of lung cancer risk in Chinese, but their results have been inconsistent. Therefore, a meta-analysis was performed to ascertain this issue. PubMed, Springer Link, Ovid and other Chinese databases were searched to include all the relevant studies. Smoking status was categorised as 'smokers' and 'non-smokers.' The pooled odds ratios (ORs) with 95% confidence intervals (CIs) were calculated using fixed or random effect model. Subgroup analyses according to ethnicity, source of control and geographical location were also conducted. This meta-analysis identified 13 studies containing 2248 lung cases and 3079 controls. Overall, a significant association between lung cancer and the variants of CYP1A1 MspI was found among smokers (type B and type C combined vs. type A: OR = 1.89, 95% CI = 1.15-3.11, P = 0.000 for heterogeneity), whereas not found among non-smokers. Similar to the overall results, stratified analyses showed that the increased risk of lung cancer was observed in population-based studies and north China among smokers (OR = 1.65, 95%CI = 1.03-2.66; OR = 2.00, 95% CI = 1.14-3.53). Our meta-analysis showed that there was an interaction between the CYP1A1 MspI and smoking on the risk of lung cancer in the Chinese population.