The current study aimed to investigate the relations of three single nucleotide polymorphisms of matrix metalloproteinase-9 gene, and single nucleotide polymorphisms-smoking interaction to subarachnoid hemorrhage risk. The optimal pattern of the interaction among single nucleotide polymorphisms and smoking was selected by generalized multifactor dimensionality reduction. The association between the three single nucleotide polymorphisms within the matrix metalloproteinase-9 gene was analyzed by logistic regression test. As well as genetic risk of subarachnoid hemorrhage interactions with smoking, the risk of subarachnoid hemorrhage in carriers with the rs3918242 (T) was significantly higher than in carriers carrying CC (genotype: CT + TT vs. CC), adjusted OR (95% CI) = 1.58 (1.25-2.03), and in carriers carrying rs17576- (genotype: AG + GG vs. AA), adjust OR (95% CI) = 1.62 (1.19-2.13). However, after adjusting for covariates, we did not find any direct association between rs17577 and subarachnoid hemorrhage risk. The generalized multifactor dimensionality reduction model shows a potential relation between rs3918242 and smoking risk for subarachnoid hemorrhage ( P = 0.0010). After covariates adjustment, current smokers with rs3918242-CT or TT genotype, compared to never-smokers with rs3918242-CC genotype, OR (95% CI) = 2.57 (1.74-3.46), have a higher subarachnoid hemorrhage risk. Our study showed that the rs3918242 (T) and rs17576 (G), the cross reaction between rs3918242 and smoking increased the risk of subarachnoid hemorrhage. Impact statement Matrix metalloproteinase-9 ( MMP-9) is a possible candidate gene for some diseases, including metabolic syndrome, stroke, coronary artery disease (CAD). But to date, limited data focused on the relationship between MMP-9 gene SNPs and SAH susceptibility. The purpose of this study was to evaluate SNPs of MMP-9 gene and their interaction with environmental factors with SAH risk based on a Chinese population.