A new procedure is described for sequencing of peptide mixtures by a combination of Edman degradation and field-desorption mass spectrometry. The procedure involves measurement of the mass values of quasi-molecular ions ([M + HI+) of constituent peptides in the mixture and their fragments degraded by the Edman method. Calculation of all the possible mass differences of the mass values before and after degradation and identification of the phenylthiohydantoins released reveal the N-terminal amino acids of individual peptides in the mixture. Repetition of these operations gives the amino acid sequences of individual peptides in the mixture. As typical examples, the sequencing of peptide mixtures prepared by proteolytic digestion of glucagon are described. In addition, a computer program was designed for determining the possible sequences of proteins from the partial sequences and mass values of their proteolytic peptides. Output data showed that the entire amino acid sequence of glucagon can be determined by only four and two cycles of degradation of chymotryptic and tryptic peptides, respectively.Analytical techniques should preferably be simple, rapid and reliable. For amino acid sequencing of polypeptides and proteins the automated sequencer technique developed by Edman and Begg [I] and gas chromatography/mass spectrometry developed by Biemann et al. [2] both have these good points. Nowadays, the former method can be successfully used for the determination of the partial or even almost complete sequences of polypeptides and proteins by cleavage of an amino acid residue from the N terminus in each degradation reaction. In contrast to the Edman procedure, gas chromatography/mass spectrometry [2,3] can be used for analysis of peptide mixtures prepared from polypeptides or proteins by partial hydrolysis, since the technique permits separation of the components and their identification in a single experiment. Although such prominent and ingenious procedures are available, more improved or different techniques are still desirable for the following reasons : first, it is desirable for chemical reactions to be as simple as possible and identification of the resulting components to be as accurate as possible; second, cross-checks by different methods are more desirable carboxypeptidase A (EC 3.4.17.1). Thus, we can determine the N-terminal amino acid residue for one given peptide in a mixture and similarly the N-terminal amino acid residues for the other components. Repeating the sequence of these operations, we can determine the amino acid sequences of