Clozapine is the gold standard and 'last resort' in treatment of refractory schizophrenia. It is important to know whether a trial with clozapine is adequate or not. Six studies show a significantly higher response rate at clozapine plasma trough levels above a therapeutic threshold of 350-400 micro g/L. The absolute risk reduction is about 40%. An additional study found best results with plasma levels between 300 and 700 micro g/L, and one (probably too small) study could not detect a significantly different response rate for 350-450 micro g/L in comparison to 200-300 micro g/L. In addition, two extension studies showed conversion from nonresponders to responders if plasma levels increased above the threshold. Investigations on time to response in treatment with clozapine are often hampered by not controlling for time until plateau of dose titration or clozapine concentration. One of the better studies found 34 responders within 8 weeks after the last dose escalation. The remaining 16 non-responding patients did not change their status during a mean follow-up of 75 weeks. A second 1 year trial found a superior differential response rate for clozapine in comparison to haloperidol only during the first 6 weeks. A third study combined regular clozapine plasma level assays with assessment of response status. At the time of response 17 out of 19 responders showed clozapine concentrations above 350 micro g/L. The nine non-responders remained below this threshold throughout the rest of the year. These results favour an approach of raising the clozapine plasma level in treatment-refractory schizophrenic patients who do not respond to an initial low-to-medium dose treatment with clozapine. Some patients, especially young male smokers, will need dosages higher than 900 mg/day. Addition of low-dose fluvoxamine while closely monitoring clozapine levels can help decrease the high number of necessary pills. An adequate trial with clozapine should last at least 8 weeks on a plasma trough level above 350-400 micro g/L.