The aim of this study was to investigate treatment failure (TF) in hospitalised community-acquired pneumonia (CAP) patients with regard to initial antibiotic treatment and economic impact.CAP patients were included in two open, prospective multicentre studies assessing the direct costs for in-patient treatment. Patients received treatment either with moxifloxacin (MFX) or a nonstandardised antibiotic therapy. Any change in antibiotic therapy after >72h of treatment to a broadened antibiotic spectrum was considered as TF.Overall, 1,236 patients (mean±sd age 69.6±16.8 yrs, 691 (55.9%) male) were included. TF occurred in 197 (15.9%) subjects and led to longer hospital stay (15.4±7.3 days versus 9.8±4.2 days; p<0.001) and increased median treatment costs (€2,206 versus €1,284; p<0.001). 596 (48.2%) patients received MFX and witnessed less TF (10.9% versus 20.6%; p<0.001). After controlling for confounders in multivariate analysis, adjusted risk of TF was clearly reduced in MFX as compared with ?-lactam monotherapy (adjusted OR for MFX 0.43, 95% CI 0.27–0.68) and was more comparable with a ?-lactam plus macrolide combination (BLM) (OR 0.68, 95% CI 0.38–1.21).In hospitalised CAP, TF is frequent and leads to prolonged hospital stay and increased treatment costs. Initial treatment with MFX or BLM is a possible strategy to prevent TF, and may thus reduce treatment costs.