A strategy for determination of the specificity of a soluble peptide. Random collections of all the possible oligopeptides of a certain length, preceded by all the possible dipeptide sequences, are prepared; for example, if hexamers, there will be 400 (20²) sets, each with the two known N-terminal residues followed by a random tetrapeptide; each of the 400 sets will consist of 160000 (20⁴) hexapeptides. Each of the 400 is tested in an assay or bioassay, e.g. binding, inhibition, vasoconstriction. The successful set is refined by placing at the first of the random positions each of the amino acids, so there will be 20 sets of hexapeptides, each with a known tripeptide sequence followed by random tripeptide sequences. These are again tested in the assay system to further narrow the selection. Eventually the most favored hexapeptide sequence is identified. A related strategy is positional scanning SPCL, in which a single amino acid is placed in each of the positions of the oligopeptide, a hexapeptide in the example, with each of the other five positions being filled at random by other amino acids. Each of these sub-libraries is assayed for selection and further refinement as in the original strategy.