The alignment of elastin molecules in the mature elastic fiber was investigated by purifying and sequencing cross-link-containing peptides generated by proteolytic digestion of incompletely cross-linked insoluble elastin. Peptides of interest were purified by reverse phase and size exclusion high performance liquid chromatography and characterized by amino acid analysis and protein sequencing. One peptide, consisting of the cross- linking domain encoded by exon 10, contained a modified lysine residue that had not condensed to form a polyfunctional cross-link. Although this domain contains the characteristic paired lysine residues found in other cross- linking domains of elastin, protein sequence analysis indicated that the first but not the second lysine had been oxidized by lysyl oxidase. This finding suggests that lysine residues in an individual cross-linking domain may not have equal susceptibility to oxidation by lysyl oxidase. In a second peptide, we found that a major cross-linking site in elastin is formed through the association of sequences encoded by exons 10, 19, and 25 and that the three chains are joined together by one desmosine and two lysinonorleucine cross-links. Past structural studies and computer modeling predict that domains 19 and 25 are linked by a desmosine cross-link, while domain 10 bridges domains 19 and 25 through lysinonorleucine cross-links. These findings, together with the high degree of sequence conservation for these three domains, suggest an important function for these regions of the molecule, possibly nucleating the aggregation and polymerization of tropoelastin monomers in the developing elastic fiber.