Introduction: Proton pump inhibitors (PPI) are one of the most prescribed medications in the world with proven efficacy. However, several studies showed that their use often doesn't respect indications, leading to over-consumption, thus exposing patients to drug interactions and adverse events (for example pneumonias). Interruption of PPIs can induce a rebound phenomenon. This generates costs for health systems.Methods: This is a prospective interventional study performed in two hospitals: La Chaux-de-Fonds (CDF, cases) and Neucha^tel (NE, control) during two six-month periods, comparing use of PPIs before and after intervention. We elaborated recommendations (PPI doses and treatment duration) based on recent medical literature that we summarized on A6 cards and gave out to all prescribing doctors in the hospital of CDF and held a 30-minute information session for the departments of surgery, medicine and anesthesiology in March 2010. Doctors were asked to apply our recommendations as often as possible, leaving space for their own assessment. No information was given to the doctors of the control hospital. The number of PPI tablets that the pharmacy sent to each careunit in both hospitals was counted and adjusted to the number of patientdays from April to September 2009 (before intervention) and April to September 2010 (after intervention). The number of other antacids that were used in both hospitals was counted during the same periods. General practitioners (GP) in the region around CDF received an explanation letter to avoid re-introduction, after discharge from the hospital, of PPI treatment stopped during the stay. The number of gastro-duodenal ulcers and upper digestive hemorrhages was counted from April to December 2009 and the same period in 2010 in both hospitals.Results: In 2010, in the hospital of CDF, the use of PPIs per 100 patient-days decreased by 36% in the surgical and medical departments compared to 2009. In the control hospital the use of PPIs per 100 patient-days increased by 10% in the surgical department and decreased by 5% in the medical department during the same periods. The decrease from 2009 to 2010 of PPI utilization in CDF comparing to NE is statistically significant: p<0.0001. Use of other antacids didn't change, ulcers or digestive hemorrhages decreased slightly from 2009 to 2010 in both hospitals. Conclusions: The study showed that with a very low-cost intervention, it is possible to decrease considerably the use of PPIs in a hospital, without taking any risk for gastro-intestinal complications.