Eight young adults who were diagnosed with autism were given the opiate antagonist naltrexone to control self-injurious behavior and maladaptive idiosyncratic mannerisms. The drug and placebo were administered in a double-blind crossover design over 17 weeks. Although one subject appeared to have partial decreases in maladaptive behaviors associated with naltrexone use, the drugs, as administered in this study, did not reduce the self-injurious and other maladaptive behaviors of the subjects.