Background: Restricted use of third-generation cephalosporins and fluoroquinolones has been linked to a reduced incidence of hospital-acquired infections with multidrug-resistant bacteria. We implemented an intensified antibiotic stewardship (ABS) programme in the medical service of a university hospital center aiming at a reduction by at least 30% in the use of these two drug classes.Methods: The ABS programme was focused on the 300-bed medical service. Prescription of third-generation cephalosporins was discouraged, whereas the use of penicillins was encouraged. Monthly drug use density was measured in WHO-ATC defined and locally recommended daily doses (DDD and RDD) per 100 patient days, to evaluate trends before (01/2008 to 10/2011) and after starting the intervention (1/2012 to 3/2013). The effect was analysed using interrupted time-series analysis with six non-intervention departments as controls.Results: Following initiation of the ABS intervention, overall antibiotic use in the medical service declined (p < 0.001). There was a significant intervention-related decrease in the use of cephalosporins and fluoroquinolones (p < 0.001) outperforming the decreasing baseline trend. Trend changes observed in some of the control departments were smaller, and the difference between trend changes in the medical service and those in control departments were highly significant for overall use and cephalosporin use reductions (p < 0.001) as well as for the increasing use of penicillins (p < 0.001). Mean use density levels (in RDD per 100 patient days) dropped for cephalosporins from 16.3 to 10.3 (-37%) and for fluoroquinolones from 17.7 to 10.1 (-43%), respectively. During the same period, the use of penicillins increased (15.4 to 18.2; 18%). The changes in expenditures for antibiotics in the medical service compared to control services minus programme costs indicated initial net cost savings likely to be associated with the programme.Conclusion: An intensified ABS programme targeting cephalosporin und fluoroquinolone use in the setting of a large academic hospital is feasible and effective. The intervention may serve as a model for other services and hospitals with a similar structure and baseline situation.