It has been suggested that the newly discovered endogenous opiate peptides (called endorphins) might play a role in the symptoms of schizophrenia. The administration of narcotic antagonists provides both a test of the hypothesis and a potential treatment. In this article, we review the methods by which data have been gathered to test endorphin involvement in schizophrenia. Alternative strategies, which hold greater promise of producing conclusive positive or negative evidence, include exploitation of individual differences, use of psychophysiological measures, genetic strategies, and multivariate statistical techniques with larger sample sizes.