Invasive mechanical ventilation can be lifesaving for patients with acute respiratory failure, but numerous complications have been identified. Therefore, once clinical improvement has occurred, emphasis is placed on quickly weaning (ie, liberating) the patient from mechanical ventilation. Weaning can be subdivided into 2 components: readiness testing and progressive withdrawal. Traditionally, both clinical factors and weaning predictors have been used to assess readiness for spontaneous breathing trials, which can be carried out using a T-piece or a low level of ventilatory support. The role of weaning predictors is under investigation, and their role in clinical decision making remains poorly defined. Recent insights into the pathophysiology of weaning failure have provided a framework for identifying potentially correctable limiting factors. Randomized controlled trials suggest that several approaches to progressive withdrawal may be acceptable, though only a minority of patients require progressive withdrawal. Emerging evidence indicates that protocol-directed weaning, driven by respiratory therapists and intensive care nurses, can improve outcome.