The constant monitoring of respiratory elastance and resistance can be of interest in patients who present a high risk of peroperative bronchospasm. The constant inspiratory flow method, proposed by Bates et al. (J Appl Physiol, 58: 1840, 1985) was chosen and automated. The inspiratory flow rate and pressure were measured respectively by a pneumotachograph linked to a differential pressure, and by a differential pressure transducer, both placed at the outlet of the inspiratory circuit. The pressure and flow signals were low-pass filtered, sampled, and then processed by an Apple II microcomputer, in order to obtain respiratory elastance and resistance. New results were displayed on the screen about every minute. The automated method was first tested in a series of 18 guinea-pigs; the respiratory parameters were compared with those obtained by the occlusion method proposed by Rossi et al. (J Appl Physiol, 58: 1849, 1985). They were found not significantly different and very strongly correlated (p less than 0.001). The ability of the constant flow method to detect changes in respiratory mechanics was then tested in a series of nine patients, after anaesthetic induction. The results obtained were in accordance with those previously published: a rise in both respiratory elastance and resistance. After giving 1 mg atropine intravenously, the respiratory resistance fell rapidly over a 5 min period, and then reached a plateau. The constant flow method, which avoids interruption in the mechanical ventilation and is sensitive to small changes in respiratory parameters, appears particularly convenient for the peroperative monitoring of patients.