Abstract Disclosure: J. Park: None. J. Kim: None. N. Kwak: None. S. Park: None. J. Yoon: None. S. Kim: None. D. Kim: None. S. Hwang: None. G. Kim: None. J. Kim: None. H. Yoo: None. J. Choi: None. Purpose: Germline copy number variation (CNV) is defined as a DNA segment with varying copy numbers resulting from duplications or deletions during evolution. The RCCX module represents one of the most complex CNV loci in humans, and its genetic variability poses challenges in accurately identifying molecular defects in the CYP21A2 gene. This study aimed to investigate the genetic heterogeneity within the RCCX module using data from the Human Pangenome Reference Consortium (HPRC), which includes whole-genome sequencing data from 47 individuals using long-read sequencing (LRS) platform. Additionally, we sought to establish an enhanced reference framework to facilitate the comprehensive genetic analysis of patients with congenital adrenal hyperplasia. Methods: To identify gene structures within the RCCX locus in the HPRC, known gene sequences from the GRCh38 reference genome were used as queries in a BLAST analysis against the genome assemblies of 94 alleles. This analysis determined the positions and sequences of these genes within each assembly, revealing diverse structural forms, including mono-, bi-, tri-, CYP21 genes, as well as TNX genes. The sequences of the CYP21 genes were then compared to the known sequences from the GRCh38 reference genome using graph-based pangenome-level variant calling to identify specific variants. Active and pseudogene-specific markers (58 for CYP21 genes and 11 for TNX genes) were used to classify the genes as active, pseudogene, or fusion. The presence of all markers indicated CYP21A1P, partial presence indicated a CYP21 fusion, and complete absence indicated CYP21A2. Results: The analysis identified ten different allele types within the RCCX region. The most prevalent configuration, observed in approximately 80% (75/94) of the samples, was bimodular, with 71% (67/94) containing one copy each of CYP21A2 and CYP21A1P. In the bimodule configuration, 70% (70/94) of individuals had one copy each of TNXA and TNXB, while four individuals exhibited a TNXA-TNXB fusion alongside a TNXB gene. One individual had two copies of TNXB without TNXA. Monomodular alleles, accounting for 14% (13/94), consistently featured the active CYP21A2 and TNXB genes. Trimodular alleles constituted 6% (6/94) with one allele containing two copies of CYP21A1P and one copy of CYP21A2, another having one copy of CYP21A1P and two copies of CYP21A2, and the remaining allele comprising two copies of CYP21A1P and one CYP21A1P/CYP21A2 fusion gene. Regarding TNX genes, three individuals had two copies of TNXA and one copy of TNXB, while another three individuals had one copy of TNXA, a TNXA-TNXB fusion, and one copy of TNXB. Conclusions: Using the LRS and genome assembly analysis, we were able to identify not only the number of CNVs but also the specific genotypes. We demonstrated the extensive allelic diversity present in the general population. Presentation: Sunday, July 13, 2025