A modification of the Edman degradation is described whereby peptides are degraded sequentially on a solid support by a machine, the solid-phase peptide sequencer. Reagents are pumped through a column containing the resin-bound peptide, and the amino acid thiazolinones released after each degradative cycle are collected on a fraction collector for subsequent analysis. A complete cycle requires 2 h, and all operations are controlled automatically by a programmer. An improved peptide resin is described, to which peptides can be attached in yields ranging from about 7001, for the 30 amino acid insulin B-chain to nearly lOOO/, for smaller peptides. Of the 18 amino acids examined so far, glycine, alanine, phenylalanine, tyrosine, valine, leucine, isoleucine, proline, methionine, asparagine, glutamine, histidine, lysine, threonine, serine and cysteic acid are degraded normally ; the phenylthiohydantoin of threonine is partially dehydrated and that of serine is almost completely destroyed by trifluoroacetic acid, however. The phenylthiohydantoin of glutamic acid is not detected, because it remains bound by its y-carboxyl group to the resin. When aspartic acid is encountered in peptides, further degradation is completely blocked, presumably as a result of the fi-carboxyl group having formed a cyclic h i d e . Using the sequencer and unpurified reagents, it was possible to degrade entirely a 0.24 pmole sample of the 21 amino acid insulin A-chain (oxidized) and to analyze the products within a 3-day period. Degradative yields averaged 94O/, per cycle. Similarly, 0.1 pmole of the insulin B-chain was degraded through 18 cycles before identification of the phenylthiohydantoins became ambiguous. The sequencer is simple in design and operation, and is inexpensive to construct. The scope and limitations of the method are discussed.The Edman degradation [l], long one of the most useful methods for amino acid sequence determination, has undergone almost continual modification and improvement over the years, culminating in the protein sequenator [2], which is capable of performing as many as 60 degradative cycles automatically. Concurrently with Edman and Begg's report [2], several laboratories [3-51 demonstrated the feasibility of degrading peptides attached to an insoluble resin, an extension of the solid-phase peptide synthesis of Merrifield [6]: The usefulness of the solid-phase method lies in its applicability to smaller peptides not adequately handled by the protein sequenator [2] and its potential for being automated.Our objective in developing the solid-phase Edman degradation has been to provide a rapid, simple and economical method for sequencing peptides containing up to about 15-20 amino acids, the size range normally found in enzymic digests of proteins. Our progress towards this goal is described here.A preliminary report of this work has been published [7].