In our opinion the best guide to prescribing antipsychotics is the clinician’s experience with his patients and in particular the patient being treated. If treatment works, stick with it. We feel it is also important for the clinician to consider the evidence from well-controlled double-blind random-assignment studies because in “evidence-based medicine,” biases both known and unknown are controlled by blinding and randomization. The purpose of this paper is to summarize and discuss the evidence on efficacy. Choice of antipsychotic, in our opinion, is probably the most important decision that the clinician makes for the psychotic patient. This involves the choice of drug, its dose, balancing efficacy, side-effects and cost.