Objective: To quantitatively assess the spontaneous breathing (SB) pattern, during minimal ventilatory support, of patients who pass or fail weaning trials from mechanical ventilation. Design: A prospective, clinical trial. Setting: Intensive care unit of a university teaching hospital. Patients: Fifty-two tracheally intubated and hemodynamically stable patients who were judged clinically ready for extubation. Methods: Using a computerized respiratory profile monitor, continuous respiratory parameters were obtained while patients were receiving four or less synchronized intermittent mandatory (SIMV) breaths and during CPAP trials. Coefficients of variation (CV) of spontaneous tidal volumes and flows during SIMV trials as well as the entropies and dimensions of the breathing patterns during CPAP trials were used to assess the dynamical breathing behaviors of the patients who passed or failed weaning trials. Measurements and results: Thirty-nine extubations were successful and 13 were not. The CV of the spontaneous tidal volumes (VT) and the spontaneous peak inspiratory flows (PF), the Kolmogorov entropy and the dimension of the SB patterns were compared in the two groups. The CV of VT (9.13±4.11 vs 26.07±6.94), the CV of PF (11.63±4.18 vs 29.88±12.07), the Kolmogorov entropy (0.09±0.03 bits/cycle vs 0.39±0.09 bits/cycle), and the dimension of the SB pattern (1.33±0.07 vs 3.93±0.47) were all significantly smaller (P<0.05) in the successfully extubated group versus the group that failed extubation. Conclusion: The spontaneous breathing pattern during minimal mechanical ventilatory support is more chaotic in patients who failed extubation trials compared to patients who passed extubation trials. Thus, we speculate that characterizing the SB pattern during minimal ventilatory support might be a useful tool in differentiating between extubation success and failure.