To investigate a modified weaning procedure to predict extubation outcome in critically older and ventilated patients.We retrospectively analyzed extubation outcome in older (≥70 years) and ventilated patients. In period I (2007), patients passing a 2-hour spontaneous breathing trial (SBT) were extubated. In period II (2008), patients underwent an 8-hour SBT on day 1 and a 2-hour SBT, followed by extubation on day 2. Weaning parameters were recorded at baseline (T 0 ) (periods I and II), 2 and 8 (T 8 ) hours after SBT (period II).The demographic data of patients in each period (n = 64 and 67, respectively) were similar. Patients in period II demonstrated a higher rate of SBT failure but a significantly lower rate of extubation failure and reintubation mortality. In period II, successfully extubated patients demonstrated a significantly lower value of rapid shallow breathing index (RSBI) at T 8 . The ratio of RSBI at T 8 over T 0 (T 8 /T 0 ≤ 1.4) demonstrated good diagnostic value (sensitivity 89.5%, specificity 80.0%, accuracy 88.4%) in predicting successful extubation.For critically older and ventilated patients, a prolonged SBT in conjunction with evolution of the RSBI ratio over baseline during SBT may serve as a useful procedure to predict extubation outcome.