Four studies using identical protocols evaluated the safety and efficacy of four novel, evidence-based targets for antipsychotic agents: a neurokinin (NK₃) antagonist (SR142801), a serotonin 2A/2C (5-HT_2A/2C) antagonist (SR46349B), a central cannabinoid (CB₁) antagonist (SR141716), and a neurotensin (NTS₁) antagonist (SR48692). Adults with schizophrenia or schizoaffective disorder (N=481) were randomly assigned in a 3:1:1 ratio to receive fixed doses of investigational drug, placebo, or haloperidol for 6 weeks. Primary efficacy variables included changes from baseline in total score on the Positive and Negative Syndrome Scale, severity of illness score on the Clinical Global Impression (CGI), and total score and psychosis cluster score on the Brief Psychiatric Rating Scale (BPRS). Significantly greater improvement in all primary efficacy variables was seen in the group receiving haloperidol than in the group receiving placebo at 6 weeks (endpoint analyses), indicating the validity of the study. The group receiving the NK₃ antagonist showed significantly greater improvement over baseline than the group receiving placebo as measured by Positive and Negative Syndrome Scale total score, CGI severity of illness score, and BPRS psychosis cluster score. Reductions in the Positive and Negative Syndrome Scale total and negative scores in the group receiving the 5-HT_2A/2C antagonist were significantly larger than those in the group receiving placebo. The improvements in psychopathology produced by the NK₃ and 5-HT_2A/2C antagonists were smaller than those produced by haloperidol, although the response to the NK₃ antagonist was positively correlated with plasma levels. The groups receiving the CB₁ and NTS₁ antagonists did not differ from the group receiving placebo on any outcome measure. All investigational drugs were well tolerated. The novel design used in this study permitted the use of a smaller number of patients receiving placebo to test the efficacy of the four novel compounds. The NK₃ and 5-HT_2A/2C antagonists showed evidence of efficacy in the treatment of schizophrenia and schizoaffective disorder. Study limitations preclude a definitive conclusion on the efficacy of CB₁ and NTS₁ antagonists in the treatment of schizophrenia. Further study of these two promising nondopaminergic mechanisms to treat schizophrenia and schizoaffective disorder appears indicated.