Steroid-sensitive nephrotic syndrome (SSNS) accounts for >80% of cases of nephrotic syndrome in childhood. However, the etiology and pathogenesis of SSNS remain obscure. Hypothesizing that coding variation may underlie SSNS risk, we conducted an exome array association study of SSNS. We enrolled a discovery set of 363 persons (214 South Asian children with SSNS and 149 controls) and genotyped them using the Illumina HumanExome Beadchip. Four common single nucleotide polymorphisms (SNPs) in HLA-DQA1 and HLA-DQB1 (rs1129740, rs9273349, rs1071630, and rs1140343) were significantly associated with SSNS at or near the Bonferroni-adjusted P value for the number of single variants that were tested (odds ratio, 2.11; 95% confidence interval, 1.56 to 2.86; P=1.68×10(-6) (Fisher exact test). Two of these SNPs-the missense variants C34Y (rs1129740) and F41S (rs1071630) in HLA-DQA1-were replicated in an independent cohort of children of white European ancestry with SSNS (100 cases and ≤589 controls; P=1.42×10(-17)). In the rare variant gene set-based analysis, the best signal was found in PLCG2 (P=7.825×10(-5)). In conclusion, this exome array study identified HLA-DQA1 and PLCG2 missense coding variants as candidate loci for SSNS. The finding of a MHC class II locus underlying SSNS risk suggests a major role for immune response in the pathogenesis of SSNS.