OBJECTIVE: The novel antipsychotics in comparison with old ones turned out to have similar clinical efficacy, produce less adverse effects, increase quality of life, reduce hospital stay with subsequent shift in resources towards community care. However, in Poland drug costs result in substantial percentage of direct healthcare costs and cost-effectiveness of novel antipsychotics can be questionable. Decision analysis model was used to evaluate potential clinical and economical consequences of using orally risperidone versus haloperidol in chronic schizophrenic Polish patients. METHODS: Decision analysis model based on a 3 month Markov cycle tree was implemented through time horizon of 5 years. The probability parameters for model (relapse rates, drop-out rates, switching to other antipsychotics, adverse effects and other transition probabilities) were obtained from literature. The clinical management pattern and cost of therapy in various model states reflected Polish schizophrenic patients treatment practice. The model was directed by clinical guidelines issued in 1997 and adjusted to current clinical practice. The costs, calculated from the sum of the charges applicable to each of the management situations over time, were expressed in 2000’PPP USD values. Bootstrapping technique was used to evaluate the 95%CI for mean cost of therapy with both alternative drugs. RESULTS: The model revealed 11% higher percentage of risperidone patients with no relapse as compared with haloperidol patients. The mean cost of therapy with risperidone equaled 11412 USD. First line treatment with haloperidol resulted in average cost of 11154 USD. The difference between both therapies tested by bootstrapping techniques was found not statistically significant. CONCLUSION: The therapy costs with risperidone and haloperidol are comparable in Polish clinical and economic settings. Higher clinical effectiveness of risperidone produces savings balancing drug cost.