OBJECTIVE Antibiotic de-escalation is part of an antibiotic stewardship strategy to achieve adequate therapy for infections while avoiding the prolonged use of broad-spectrum antibiotics. However, there is a paucity of clinical evidence on the clinical impact of this strategy in pneumonia patients in the intensive care unit (ICU). This review aimed to evaluate the impact of antibiotic de-escalation therapy for adult patients diagnosed with pneumonia in the ICU. This review was conducted in accordance with the Meta-analysis of Observational Studies in Epidemiology (MOOSE) recommendation. Electronic databases including MEDLINE, CINAHL, PubMed, Embase, Cochrane Databases and Cochrane Central Register of Controlled Trials were searched up to March 2017 for relevant trials. The methodological quality of included trials was assessed by using a modified version of the Newcastle-Ottawa Quality Assessment Scale for Case-Control and Cohort Studies. A meta-analysis was conducted using the random-effect model to combine the rate of mortality and length of stay outcomes. Nine observational trials involving 2128 patients were considered eligible for inclusion. Although based on low quality evidence, there was a statistically significant difference in favour of the impact of de-escalation on hospital stay but not mortality (MD -5.96 days; 95% CI -8.39 to -3.52). This review highlights the need for more rigorous studies to be carried out before a firm conclusion on the benefit of de-escalation therapy is supported.