Multiple DNA and protein sequence alignments have been constructed for the human T-cell receptor cVS, [3, and y (TCRA/D, B, and G) variable (V) gene segments. The traditional classification into subfamilies was confirmed using a much larger pool of sequences. For each sequence, a name was derived which complies with the standard nomenclature. The traditional numbering of V gene segments in the order of their discovery was continued and changed when in conflict with names of other segments_ By discriminating between alleles at the same locus versus genes from different loci, we were able to reduce the number of more than 150 different TCRBV sequences in the database to a repertoire of only 47 functional TCRBV gene segments. An extension of this analysis to the over 1 O0 TCRAV sequences results in a predicted repertoire of 42 functional TCRAV gene segments. Our alignment revealed two residues that distinguish between the highly homologous V8 and Vo~ one at a site that in VH contacts the constant region, the other at the interface between immunoglobulin Vn and VL. This site may be responsible for restricted pairing between certain V8 and Vy chains. On the other hand, V[3 and V 7 appear to be related by the fact that their CDR2 length is increased by four residues as compared with that of VcV8 peptides.The alignment data reported in this paper have been submitted to the EMBL nucleotide sequence database and have been assigned the alignment number DS23485. The data are available by the EBI FTP server and file server