OBJECTIVE: Pharmacoeconomic evaluation of selective serotonin reuptake inhibitors (SSRIs) is becoming an important issue for regulatory and marketing purposes. At the time of the launch of the drug, decision-tree models (i.e. Markov and Monte-Carlo) provide the most realistic projection data available. Designed to provide a large range of information on outcome and cost, which are based on various assumptions, the models enable decision-makers (physician, pharmacist, HMO, regulatory etc.) to make informed, relative decisions. METHODS: This poster illustrates five different model structures of associated expenses that are incurred during the treatment of Depression. The “branches” on these models become progressively complex, and are based on the following five factors: (1) severity of the depression, (2) rate of compliance, (3) success/failure rate, (4) patient dropout from the health care system (patient having no contact with any medical institution or any medical doctor), and (5) switch in case of failure or non tolerability. Costs were taken from the literature using different sources. RESULTS: SSRIs could be more or less cost-effective than tricyclic antidepressants (TCAs) depending on whether switch, compliance, success of treatment, dropout from the health care system and disease have been taken into consideration. Conversely, when the same information is derived from different sources of the same country, the cost-effectiveness ratio is significantly affected. Moreover, sensitivity analysis does not prevent such differences in outcome. CONCLUSIONS: Caution should be used when reviewing such data as these results indicate the tendency for gross discrepancy to exist between different methods of analysis. These results thus emphasise the need for standardised pharmacoeconomic models. The authors make arguments in support of this endeavour.