Kallikreins are serine proteases that release kinins from kininogens. Kinins, via their effects on cardiovascular and renal function, may be involved in the pathogenesis of hypertension and renal failure. Two groups of kallikreins exist, glandular or tissue kallikrein and plasma kallikrein. In this study, we examined the human plasma kallikrein gene KLK3 to determine whether it contributed to end-stage renal disease (ESRD) susceptibility. We identified two novel polymorphic sequences closely linked to the KLK3 gene, designated KLK3b and KLK3c (heterozygosities: 0.64 to 0.68 and 0.48 to 0.52, respectively). We mapped the KLK3 gene and the marker KLK3c to the long arm of human chromosome 4 between F11 and D4S426 using a radiation hybrid panel. The study population consisted of 142 sibling pairs concordant for ESRD from 121 African American families. The 142 sibling pairs were stratified into 78 pairs with hypertension- and chronic glomerulonephritis-associated ESRD and 64 with non-insulin-dependent diabetes mellitus-associated ESRD. Linkage analyses, using SIBPAL of SAGE, and exclusion analysis, using MAPMAKERS/SIBS, were performed. Linkage analysis of affected sibling pairs did not reveal any evidence of linkage of KLK3 to ESRD in all 142 sib-pairs or in the two stratified subsets. Exclusion analysis indicated that the KLK3 gene could be excluded from contributing to ESRD at a relative risk of 3 when the maximum log of the odds score of -2 was used as the criterion for exclusion. However, an association analysis using the relative predispositional effect technique showed that alleles 7 and 9 of KLK3b were consistently associated with ESRD. Alleles 7 and 9 were present in 11.2% and 10.8% of the 113 unrelated ESRD probands and in 6.6% and 6.6% of the 204 race-matched control subjects without renal disease (allele P=.0041 and .0016, respectively). Alleles 7 and 9 were also present in 13% and 10.4% of the proband's first siblings (allele P=.00014 and .0087, respectively). The association of KLK3b alleles with ESRD raises the possibility that polymorphisms in KLK3 may play a role in ESRD susceptibility. The lack of linkage might reflect our relatively small family set.