Neuropeptides containing the carboxylterminal sequence Arg-Phe-NH2 are found throughout the animal kingdom and are important substances mediating neuronal communication. Here, we have cloned the cDNA coding for the precursor protein of the sea anemone neuropeptide (Antho-RFamide) less than Glu-Gly-Arg-Phe-NH2. This precursor is 334 amino acids in length and contains 19 copies of unprocessed Antho-RFamide (Gln-Gly-Arg-Phe-Gly), which are tandemly arranged in the C-terminal part of the protein. Paired basic residues (Lys-Arg) or single basic residues (Arg) occur at the C-terminal side of each Antho-RFamide sequence. These are likely signals for posttranslational cleavage. The processing signals at the N-terminal side of each Antho-RFamide sequence, however, include acidic residues. Processing at these amino acids must involve either an amino- or an endopeptidase that cleaves C-terminally of aspartic acid or glutamic acid residues. Such processing is, to our knowledge, hitherto unknown for peptidergic neurons. The Antho-RFamide precursor also contains two copies of the putative Antho-RFamide-related peptide Phe-Gln-Gly-Arg-Phe-NH2 and one copy of Tyr-Val-Pro-Gly-Arg-Tyr-NH2. In addition, the precursor protein harbors four other putative neuropeptides that are much less related to Antho-RFamide. This report shows that the biosynthetic machinery for neuropeptides in coelenterates, the lowest animal group having a nervous system, is already very efficient and similar to that of higher invertebrates, such as mollusks and insects, and vertebrates.