Diabetes mellitus is a multifactorial disorder characterized by a high level of glucose in the blood. Both genetic and environmental factors interact to cause diabetes. Insulin receptor substrate (IRS) proteins have a significant part in insulin signaling pathways. We aimed to investigate the relationship of type 2 diabetes with a Gly972Arg (G972R) variant of the IRS-1 gene and Gly1057Asp (G1057D) polymorphism of IRS-2 gene in the population of Punjab, Pakistan.We collected 926 samples, 500 healthy controls (fasting blood sugar  99 mg/dL, random blood sugar > 126 mg/dL). Several anthropometric measurements were measured. Statistical analysis was performed by using SPSS to determine the allele group/genotype frequency of the selected variants in the study population.The genotyping results of G972R by RLFP-PCR showed the allelic frequency of G = 0.68 and R = 0.32 in controls while G = 0.71 and R = 0.29 in the cases. The minor R allele had a slightly higher frequency in the cases than the controls (OR = 0.86, CI 0.706-1.052, p = 0.17). The genotyping results of G1057D showed allelic frequency G = 0.74 and D = 0.26 in the controls while G = 0.961 and D = 0.29 in the cases. The minor D allele appeared to be a risk allele for this SNP although the difference in the allele frequencies was not statistically significant (OR = 1.55, CI 0.961-1.41, p = 0.108). The combined genotype analysis showed that the difference in the allele and genotype frequencies reached statistical difference between the cases and the controls as well as the odds ratio substantially increased when the R allele (G972R) was present together with D allele (G1057D) in any combination. When the association of single variants with the lipid traits was observed, only D allele (G1057D) showed significant association with TG, HDL and LDL, however when the analysis was repeated for combined genotypes using general linear model, many more significant associations between the genotype where D allele and R allele are together, were seen with many lipid traits.In conclusion, the single nucleotide polymorphisms with low-modest effect size may not affect the phenotype individually but when in combination, the effect becomes stronger and more visible, therefore, for the SNP association studies, the more the number of SNPs included in the analysis, the more meaningful the results.