We review evidence from randomized, placebo-controlled studies of patients with schizophrenia or schizoaffective disorder, which compared 2 or more doses of an antipsychotic to calculate the dose-response curve for each first-generation (typical) antipsychotic (FGA) or second-generation (atypical) antipsychotic (SGA) and as a group (based on dose equivalence). We identified the near-maximal effective dose (ED; ie, the threshold dose necessary to produce all or almost all the clinical responses for each drug). In randomized, fixed-dose studies of SGAs, the near-maximal efficacy dose for olanzapine may be greater than 16 mg; for risperidone, it is 4 mg; and for ziprasidone, it is 120 mg. Risperidone at 2 mg daily is 50% less efficacious than higher doses. Olanzapine at about 6 mg is approximately 33% less effective than higher doses. Aripiprazole at 10 mg daily was fully efficacious. Doses of clozapine well above 400 mg are necessary for optimal treatment of many schizophrenia patients. We found 3.3 to 10 mg haloperidol to be the near-maximal ED range. We find no evidence that doses higher than these are more effective. We failed to find that high doses of haloperidol (or all other first-generation comparison drugs converted to equivalent doses) were less effective than medium doses (3.3 to 10 mg). While high-dose FGAs are not less effective, we feel it is important not to avoid using high dose to avoid excessive toxicity.