In the clinical practice, the mechanical ventilation is a very important assisting method to improve the patients' breath. Whether or not the parameters set for the ventilator are correct would affect the pulmonary gas exchange. In this study, we try to build an advisory system based on the gas exchange model for mechanical ventilation using fuzzy logic. The gas exchange mathematic model can simulate the individual patient's pulmonary gas exchange, and can help doctors to learn the patient's exact situation. With the fuzzy logic algorithm, the system can generate ventilator settings respond to individual patient, and provide advice to the doctors. It was evaluated in 10 intensive care patient cases, with mathematic models fitted to the retrospective data and then used to simulate patient response to changes in therapy. Compared to the ventilator set only as part of routine clinical care, the present system could reduce the inspired oxygen fraction, reduce the respiratory work, and improve gas exchange with the model simulated outcome.