Abstract: Using a newly constructed dataset on German hospitals, which includes 24 process and outcome indicators of clinical quality, we test whether quality has increased in various clinical areas since the introduction of mandatory quality reports and the online publication of part of the collected quality measures. Our results suggest that process indicators of clinical quality have increased significantly in 2008 compared to 2006. In addition, the hospitals underperforming in 2006 appear to have increased their clinical quality relatively more than the other hospitals. When instead quality is measured by outcome indicators, average clinical quality is estimated to have increased for underperforming hospitals and decreased for the best performing hospitals in 2006, so that on average across all hospitals the changes in outcome indicators are insignificant for just more than half of the outcome quality measures. We further show that the best performing hospitals in 2006 in terms of outcome quality measures experienced an increase in their share of patients in 2008, thus providing indirect evidence that patients react to disclosed quality. Interestingly, the best performing hospitals in 2006 in terms of process quality measures did not experience a significant change in their share of patients in 2008, thus suggesting that patients react more to output than to process measures of quality. Finally, for the subset of hospitals who offer services in obstetrics, we find that higher competitive pressure, measured as the number of competitors in a given radius, is associated with a higher increase in quality following quality disclosure. We argue that the latter effect is unlikely to be due to selection of patients by hospitals.