Abstract Ventilator-associated pneumonia (VAP) caused by multidrug-resistant (MDR) Gram-negative bacteria (GNB) has emerged as an important and intractable clinical problem. This review assessed the efficacy and safety of colistin for treatment of MDR GNB VAP. PubMed and Embase were searched for controlled studies of colistin for treatment of MDR GNB VAP. The Mantel–Haenszel random-effects model was used to pool odds ratios (ORs) with 95% confidence intervals (CIs). The primary outcome was clinical cure; secondary outcomes were microbiological eradication, ICU mortality, hospital mortality, length of ICU stay and nephrotoxicity. Fourteen controlled studies involving 1167 patients were identified, including six reporting colistin versus β-lactam antibiotics, three reporting aerosolised (AS) plus intravenous (IV) colistin versus IV colistin alone and five reporting colistin combined therapy versus colistin monotherapy. The clinical cure rate of colistin was comparable with that of β-lactam antibiotics (OR=1.00, 95% CI 0.68–1.47). Compared with IV colistin alone, AS plus IV colistin exhibited a better clinical cure (OR=2.12, 95% CI 1.40–3.20). Compared with colistin monotherapy, colistin combined therapy did not appear to provide a better clinical cure (OR=1.38, 95% CI 0.81–2.33). There was no significant difference in nephrotoxicity and other secondary outcomes between the treatment groups. Colistin appears as effective and safe as β-lactam antibiotics for the treatment of MDR GNB VAP. AS colistin may be a beneficial adjunct to IV colistin in the management of MDR GNB VAP. Colistin combined therapy does not appear to provide better outcomes compared with colistin monotherapy.