CD2 (T11, sheep erythrocyte receptor) is a surface antigen of the human T-lymphocyte lineage. cDNA clones encoding CD2 have been isolated by using the purified, denatured CD2 to raise a rat antiserum. Positive clones were recognized in a phage lambda gt11 expression library prepared from the human leukemia T-cell line J6. The DNA sequence contained an open reading frame encoding 360 amino acids. The N-terminal 24 amino acids were characteristic of a signal peptide and were followed by a region that matched all 25 residues of the CD2 N terminus previously determined by amino acid sequencing. The predicted amino acid sequence is consistent with that of a transmembrane glycoprotein containing three potential N-glycosylation sites on the N-terminal side of a 26-amino acid hydrophobic segment. There is a large cytoplasmic domain of 125 amino acids that is rich in proline and in basic residues. RNA blot-hybridization analysis demonstrated hybridization only in those T cells that were positive for surface CD2 antigen. There are limited regions of sequence similarity to members of the immunoglobulin supergene family.