This article raises questions about the morality and value of experiments conducted mainly on psychiatric patient-subjects whose mental capacity and judgment are often impaired, making them incapable of giving informed consent. Its focus is on experimental studies in which psychotic symptoms in patients with schizophrenia have been knowingly exacerbated by suddenly withdrawing medications that they needed, administering known psychosis-producing substance such as L-dopa and apomorphine, and ignoring the treatment needs of those serving as experimental controls in placebo studies. Concerns are raised about the draft “Statement of Principles for Ethical Conduct” by the American College of Neuropsychopharmacology. Questions are also raised about the adequancy of current safeguards, including federal regulations, peer review, and the trivialization of “informed consent” by institutional review boards that operate under veils of secrecy. Implications for mental health policy are discussed, and suggestions are made for improving safeguards and reducing risks.