OBJECTIVES: The aim was to compare prescription patterns in Poland in treatment of community-acquired pneumonia (CAP), matched with current guidelines and their financial consequences. METHODS: Epidemiological data from Alexander Project microbiological study were used. The Dataview Medical Database IMS Health for Poland, was the source of medical data on provided medical services and prescriptions made out by out-patient clinic doctors. These were compared with guidelines of Polish Working Group for Standards of Rational Therapy and Prophylaxis of Infections. Drugs prices were taken from offer of leading wholesaler. RESULTS: CAP makes up of 4% of all respiratory system infections, 20% of them require hospitalisation. The Alexander Project shows that causes of CAP are: S. pneumoniae (7-76%), H. influenzae (1-16%), S. aureus (0-4%), Gram (-) bacteria (0-28%), non-typical bacteria such M. pneumoniae (0-24%) C. pneumoniae and Legionella sp.(0-15%). In 2002, for patients aged 12-64 years, 302.656 visits (234.451 - first visits, 68.206 - follow-up visits) because of CAP were registered, in 254.179 (i.e. 84%) cases prescriptions for antibiotics were made-out. The percentages of antibiotics were following: 24.1% oral macrolid, 4.4% injected macrolid, 24.2% oral cephalosporin, 16.4% injected cephalosporin, 14.4% oral wide-spectrum penicillin, 11.6% oral tetracycline, 10.2% aminoglycoside, 2.2% oral fluoroquinoline, 0.2% co-trimoxasole (combined treatment 7.7%). Total cost of antibiotics was €4.549 million. For first visits this cost amounted to €4.043 million and for follow-up visits to €0.507 million. The cost of treatment according to the guidelines was estimated, depending on chosen option, at €3.260 to €6.713 million. CONCLUSIONS: Analysis shows that prescription patterns in CAP, differs from the guidelines. 45% of cases were treated in way foreseen for non-typical CAP, whereas epidemiological data shows that real number of such cases does not exceed 20%. Treatment based on recommendations would not bring savings, but would slow down the growth of bacterial resistance.