OBJECTIVE. To assess the economic consequences of changing the prescription pattern of proton pump inhibitors (PPI) in Sweden by use of a budget impact model (BIM). METHODS. Data from published sources provided information about treatment effects of different PPIs and productivity losses due to upper gastrointestinal symptoms (GIS). Sale statistics at the county council level and official Swedish price lists were used to estimate drug costs. Data on wages ware taken from Statistics Sweden and was used to estimate indirect costs. The BIM was flexible and allowed us to analyse (i) the drug budget impact in a particular county council of switching between different PPIs and (ii) the annual per patient drug cost and the indirect cost from treating patients with different PPIs during a specific time period. RESULTS. In the county council of Stockholm (1.9 million inhabitants) the annual PPI drug costs amount to about €11.5 Million of which 50% and 14% comprise of generic omeprazole and esomeprazole, respectively. The consequence of switching from generic omeprazole to esomeprazole for 30 % of the patients would result in increasing the PPI drug costs by 11 %. However, as the major part of the annual per patient costs comprise of indirect costs due to productivity losses and absence from work, esomeprazole only need to be 1-2 percent units more effective in reducing GIS per week than generic omeprazole to compensate for its higher price. CONCLUSION. The results indicate the importance of including both direct and indirect costs in BIMs when analysing changes in drug prescription patterns.