Background: An increasing number of older patients are prescribed proton pump inhibitors (PPIs). However, the extent of inappropriate PPI prescribing in this group is largely unknown. Objective: We sought to identify clinical and demographic factors associated with inappropriate PPI prescribing in older patients and to assess the effects of a targeted educational strategy in a controlled hospital environment. Methods: Clinical and demographic characteristics and full medication exposure on admission were recorded in 440 consecutive older patients (mean -SD age 84 - 7 years) admitted to a teaching hospital between 1 February 2011 and 30 June 2011. A 4-week educational strategy to reduce inappropriate PPI prescribing during hospital stay, either by stopping or reducing PPI doses, was conducted within the study period. The main outcome measures of the study were the incidence of inappropriate PPI prescribing and the effects of interventions to reduce it. Results: On admission, PPIs were established therapy in 164 patients (37%). This was considered inappropriate in 100 patients (61%). Lower Charlson Comorbidity Index score (odds ratio [OR] 0.76; 95% CI 0.57, 0.94; p = 0.006) and history of dementia (OR 1.65; 95% CI 1.28, 1.83; p = 0.005) were independently associated with inappropriate PPI prescribing. Interventions to reduce inappropriate PPI prescribing occurred more frequently during and after the education phase (frequency of interventions in patients with inappropriate PPI prescribing: pre-education phase 9%, during education phase 43%, and post-education phase 46%, p= 0.006). Prescribing interventions were not associated with acid rebound symptoms. Conclusions: Inappropriate PPI prescribing in older patients is frequent and independently associated with co-morbidities and dementia. A targeted inhospital educational strategy can significantly and safely reduce inappropriate PPI prescribing in the short term.