We have devised a method for assessing the relatedness of amino acid interchanges in the variable regions of immunoglobulin polypeptide chains, in terms of the nature of the differences between trinucleotide codons. We have used it to construct a map of the average human κ chain showing the positions which (i) give the chain κ identity, (ii) give the chain subgroup identity within the κ type, (iii) give the chain identity within a subgroup, and (iv) are invariant. In order to determine the positions at which the interchanges are homologous, we have performed a second analysis using the assumption that functionally equivalent amino acids are identical.κ Identity was found throughout the amino-terminal half of the chain (positions 1–108, the variable region). Subgroup identity was not found beyond invariant Phe 98, but was very pronounced in the first 22 positions of the chain. κ and subgroup identity were evident within hypervariable regions (positions 27–35, 48–52, and 90–97), however. A high degree of functional homology was found in the 11 residues at the end of the variable region and at certain individual positions scattered throughout the variable region, but not in the hypervariable regions.The distribution of these different classes of positions might reflect the distribution of certain functions within the chains.