This retrospective study examined adjunct divalproex (N = 15) or lithium (N = 9) in treatment-resistant schizophrenia patients added to clozapine and compared to clozapine monotherapy (N = 25). Six month total BPRS scores were similarly improved in all treatment groups, however significantly greater improvements occurred in the first month for those on divalproex (-9) or lithium (-8) vs. clozapine alone (-4.5) (F = 3.32, df = 10.43, p = 0.0003). Rates of sedation, tachycardia, orthostasis, GI disturbances, confusion and dizziness were similar among groups. Mean weight gain was 8.7 pounds for clozapine monotherapy, 3.0 pounds in the adjunct divalproex group and 13.3 pounds in the adjunct lithium group (P = NS). A trend was noted for greater increases in blood glucose levels for those treated with adjunct lithium (F = 2.62, df = 2.28, p = 0.09). The addition of divalproex was significantly more effective in reducing global symptoms (driven by hostility and anxiety) in the first month of adjunct treatment as compared to clozapine monotherapy and to previous clozapine treatment.