Protection of subjects in psychiatric research is an issue of considerable public and professional interest. Perhaps the most hotly debated issue concerns challenge study protocols where symptoms of illness are increased in a bioassay designed to gain knowledge of pathophysiology. Although widely used in biomedical research, the ethics of this application in mental illness research are contested. At issue is whether acute distress and lasting harm are caused without direct benefit in vulnerable subjects without valid informed consent. The ketamine challenge study in schizophrenia subjects is at the vortex of the current debate. This report presents background data on ketamine safety and a qualitative and quantitative analysis of data from all schizophrenia subjects in North American ketamine studies. Duration and severity of change in psychosis and anxiety, "worst case" experiences, and information on prolonged adverse effects are detailed. The vulnerable population and informed consent issue is discussed. Group results show that psychosis increase is mild to moderate and brief, anxiety is mild and brief, and no evidence of prolonged adverse effects is found. Few "worst case" incidents were identified, and these were clinically managed successfully in a short time period. Although more difficult to evaluate, informed consent procedures seem adequate, and consent was voluntary in subjects judged to have decisional capacity for this purpose and in circumstances where alternative clinical care could be freely chosen. The author concludes that ketamine challenge studies meet ethical standards, have been conducted without lasting adverse effects, that discomfort is modest and brief, and important new knowledge has been gained of potential benefit to the class from which subjects were drawn.