Patient-ventilator asynchrony has been associated with adverse outcomes. The largest body of investigation has focused on ineffective ventilator triggering. Nevertheless, the effect of other patterns of asynchrony on patient outcomes is unknown. The purpose of this study was to assess the performance of specific patterns of asynchrony in their ability to predict prolonged mechanical ventilation. Patients mechanically ventilated within 48 hours of intensive care unit admission were included. Subjects with tracheostomy, mechanical ventilation dependency and treatment with neuromuscular blockers were excluded. Asynchrony patterns were collected on daily evaluations for three days. Analysed patterns were missed, double and auto-triggering, dish-out and overshoot of the pressure waveform, delayed termination and auto-positive end-expiratory pressure. Pattern-specific and composite asynchrony indices were calculated. Demographic data, Acute Physiology and Chronic Health Evaluation (APACHE) II, PaO2/FiO2, positive end-expiratory pressure (PEEP) and Richmond Agitation Sedation Scale were collected. Receiver operating characteristic curves to assess the ability of each index to predict prolonged mechanical ventilation were constructed. Twenty-eight patients were deemed eligible. The average age was 54±17, with 71% of male gender. APACHE II, PaO2/FiO2 and PEEP were 18±7, 249±117 and 5.8±2.4 respectively. Richmond Agitation Sedation Scale was -3.1±1.3. The average number of days on mechanical ventilation was 5.4±6.5. Areas under the curve (AUC) for missed and double-triggering indices were 0.66±0.12 and 0.60±0.12 respectively. AUC for the dish-out index was 0.88±0.09. AUC for overshoot, delay termination and composite indexes were 0.55±0.12, 0.62±0.12 and 0.70±0.10 respectively. Dish-out index is the best predictor of prolonged mechanical ventilation, compared with other patterns of patient-ventilator asynchrony.