To identify the incidence and outcome of extubation failure in patients with burn. Retrospective cohort study in a tertiary burn intensive care unit. A review of the casenotes of 132 consecutive adult patients with burn admitted between 2001 and 2005, and requiring mechanical ventilation for >24 hours, was undertaken. Sixty-seven patients underwent extubation and entered data analyses. Extubation failure was defined as reintubation within 48 hours. The outcomes of interest were incidence and cause of extubation failure, duration of mechanical ventilation (DMV), length of stay (LOS), and mortality. The patients who succeeded and failed extubation were similar in terms of age, sex, burn size (P = .3), and incidence of inhalation injury (P = .1). Of 67 planned extubations, 20 (30%) failed. DMV (22.5 vs. 4 days; P < .001), intensive care LOS (1.20 vs. 0.41 days/% burn; P < .001), and hospital LOS (1.90 vs. 1.18 days/ % burn; P < .003) were significantly longer in reintubated patients when compared with those who extubated successfully. Extubation outcome, burn size, and age provided the best predictive model for patient outcome (P = .02), but extubation outcome was the only predictor that operated individually (P = .05). The aetiology of extubation failure in 15 (75%) patients was poor pulmonary toilet. The incidence of extubation failure in this homogenous population of patients with burn is higher than general intensive care patients (30% vs. 4-23%). The DMV, lengths of intensive care, and hospital stay are significantly longer in patients who failed extubation. In addition to burn size and age, extubation outcome is an important predictor of intensive care mortality. The indication for reintubation in most patients is poor airway clearance.