A novel computerized program has been developed for predicting the secondary structure of proteins from their amino acid sequences. The scheme of the Chou and Fasman method (1978, Adv. Enzymol. Relat. Subj. Biochem. 47, 45-148) is closely followed. Some of their qualitative rules have been converted to numeric scales to obtain unambiguous predictions. This program has been tested on 21 proteins with known three-dimensional structures constituting a 4457 amino acids data base. The percentage of correctly predicted amino acids is between 41 and 66% for a three-state (helix, sheet, and coil) description of protein secondary structure.