An aspect of the ethics of the medication-free research debate is the question of long-lasting damage from short-term psychosis. Data from medication-free research refute the hypothesis that short-term increase in psychosis during research causes long-term deterioration. More compelling are the preneuroleptic biased cohort studies, which suggest that years of untreated psychosis increase the incidence of poor outcome cases. Comprehensive treatment involves many therapeutic elements. In regard to pharmacotherapy, it is generally the case that the more aggressive drug treatment is associated with less psychosis, but more negative symptoms, poorer medication compliance, more depression, and more adverse drug effects. These complexities must be considered in evaluating ethical and safety issues. Standards should be high, and the evidence suggests that treatment in the context of research can (and usually does) meet these standards.