Early detection and intervention is offering a compelling new perspective on the diagnosis and treatment of schizophrenia spectrum disorders. Relevant investigations include efforts to reduce the post-onset duration of untreated psychosis and efforts to identify and treat prodromally symptomatic high-risk patients in the pre-onset phase of the disorder. There appears to be consensus that the benefits outweigh the risks in studies aiming to treat first episode psychotic patients as soon as possible. In contrast, there is less consensus about pre-onset detection and intervention studies, the major concerns being the false positive rate of case identification, the evidence of benefit, the side effects of treatment, the potential harm from being informed that one is at risk for psychosis, the false negative rate of case identification, and the inclusion of adolescents. In this article, each of these issues is elaborated upon and discussed in terms of the existing evidence informing risk and benefit and in light of our own clinical experience conducting this research. It is concluded that we do not have enough data yet to justify pre-onset detection and intervention as standard practice, but we do have enough data concerning risk and benefit to justify pre-onset detection and intervention research.