To determine human Ig heavy chain variable region (V H ) gene segment organization on individual homologous chromosomes, an efficient approach has been developed. Single spermatozoa were used as subjects for the study. Upon sperm lysis, V H regions in each sperm were randomly sheared into fragments by the random Brownian force. The fragments were separated from each other by aliquoting the lysate into a certain number of tubes. The gene segments in the V H 1 and V H 4 families in each tube were identified by denaturing gradient gel electrophoresis after PCR amplification. The polymorphic V H sequences were used to determine the parental origins of the analyzed sperm. V H segment organization in the parental haplotypes was determined by aligning the overlapping fragments from the spermatozoa with the corresponding haplotypes. Based on this comparison between the resulting haplotype maps and the composite map reported previously, the V H region on chromosome 14 could be subdivided into four portions. The numbers and compositions of the V H gene segments differ considerably among the maps in two portions, but are highly conserved in the other two. The data also indicate that the V H region on chromosome 15 may contain a large duplicated block with copy number varying among haplotypes. The approach used in the present study may be used to construct high-resolution haplotype maps without molecular cloning.For many human multigene families, determination of gene organization on individual homologous chromosomes is a challenging issue for several reasons: (i) the diploidy of the human genome, (ii) the genes sharing a high degree of sequence identity in each family, (iii) variation in DNA sequences and gene compositions among the haplotypes, (iv) different chromosomal locations, (v) occupation of large chromosomal regions, and (vi) recent duplications. In this paper, we show that all of these issues can be addressed by analyzing single DNA fragments from the haploid genomes of individual spermatozoa.Two human Ig heavy chain variable region (V H ) families, V H 1 and V H 4, were used for the analysis. Human V H gene segments are located on three chromosomes, 14q32, 15q11, and 16p11 (1-6). However, only about half of the V H segments on chromosome 14 are functional (7,8). Each haploid human genome contains Ϸ120 distinct V H gene segments (5,7,8). The V H segments are classified into seven families (V H 1-V H 7) with segments sharing Ͼ80% sequence identity in each family (9-15). Extensive polymorphisms for the V H region on chromosome 14 have been reported (16)(17)(18)(19)(20)(21)(22). A composite map for this region was constructed recently (8). Several studies showed that the number and composition of the V H gene segments in some portions of the V H region on chromosome 14 varied among the haplotypes (23-30). However, the extent of variation in the entire region remains unclear because no haplotype maps for this region have been constructed. The V H segment organization on the other two chromosomes has not been ...