There is no evidence supporting the use of de-escalation therapy (DET) among patients with community-acquired pneumonia (CAP). We assessed the outcomes associated with DET among bacteraemic CAP patients. We performed a secondary analysis of the Community-Acquired Pneumonia Organization database, which contains data on 660 bacteraemic patients hospitalized because of CAP in 35 countries (2001-2013). Exclusion criteria were death within 72h from admission and an inappropriate empirical antibiotic regimen. DET was defined as changing an appropriate empirical broad-spectrum regimen to a narrower-spectrum regimen according to culture results within 7 days from hospital admission. Two study groups were identified: patients whose antibiotic therapy was de-escalated (the DET group), and patients whose antibiotic therapy was not de-escalated (the N-DET group). The primary study outcome was 30-day mortality. Two hundred and sixty-one bacteraemic CAP patients were included. Gram-positive bacteria were responsible for 88.1% of the cases (Streptococcus pneumoniae, 75.9%). Gram-negative bacteria were responsible for for 7.3% of the cases. DET was performed in 165 patients (63.2%). The N-DET group was characterized by a more severe presentation at admission. After adjustment for confounders, DET was not associated with an increased risk of 30-day mortality. DET seems to be safe among bacteraemic patients with CAP. Randomized clinical trials are warranted to further explore these findings.