Ventilator-associated pneumonia (VAP) as a result of multidrug-resistant Gram-negative bacteria has contributed to the revival of the use of intravenous (i.v.) colistin. However, the additional administration of inhaled colistin for VAP is controversial. We performed a retrospective cohort study of patients with microbiologically documented VAP who received i.v. colistin with or without inhaled colistin. Seventy-eight patients with VAP received i.v. plus inhaled colistin, whereas 43 patients received i.v. colistin alone. The mean +/- SD daily dosage of i.v. colistin was 7.0 +/- 2.4 and 6.4 +/- 2.3 million international units (IU), respectively (p 0.13); the average daily dosage of inhaled colistin was 2.1 +/- 0.9 million IU. The outcome of infection was cure for 62/78 (79.5%) patients who received i.v. plus inhaled colistin vs. 26/43 (60.5%) patients who received i.v. colistin alone (p 0.025); all-cause in-hospital mortality was 31/78 (39.7%) vs. 19/43 (44.2%), respectively (p 0.63); all-cause intensive care unit (ICU) mortality was 28/78 (35.9%) vs. 17/43 (39.5%), respectively (p 0.92). The use of inhaled colistin was independently associated with the cure of VAP in a multivariable analysis (OR 2.53, 95% CI 1.11-5.76). Independent predictors of mortality were a higher APACHE II score (OR 1.12, 95% CI 1.04-1.20), presence of malignancy (OR 4.11, 95% CI 1.18-14.23) and lower daily dosage of i.v. colistin (OR 0.81, 95% CI 0.68-0.96). The outcome of VAP was better in patients who received inhaled colistin with i.v. colistin than those who received i.v. colistin alone. There was no difference in all-cause in-hospital and ICU mortality between the two groups. Randomized controlled trials are needed to evaluate further the role of inhaled colistin in VAP.