An animal model was prepared to compare the effectiveness of high-frequency jet ventilation (HFJV) with that of conventional intermittent positive pressure ventilation (IPPV) in managing bronchopleural fistula. In ten adult beagles, the left upper pulmonary lobe was resected and the left upper lobe bronchus was cannulated to establish a permanent bronchopleural fistula. Apposition of the middle lobe to the chest wall was examined via an open thoracotomy. Fistula leakage was measured with the thoracotomy closed, and chest tube drainage was set at 20 cm H2O. With HFJV, lung apposition was attained at lower peak (8.9 cm H2O) and expiratory (5.7 cm H2O) pressures than with IPPV (21.2 cm H2O Peak pressure and 10.7 cm H2O expiratory pressure). With equivalent arterial blood gases, fistula leakage was 47 per cent higher with IPPV than with HFJV (P less than .001). Each 5 cm of positive end expiratory pressure (PEEP) added increased fistula leakage similarly to both systems. Varying the frequency of HFJV did not alter leakage significantly, but increasing the driving pressure markedly increased leakage. Thus, experimentally, HFJV proved superior to IPPV in the management of bronchopleural fistula. When HFJV is used clinically to manage a bronchopleural fistula, the driving pressure should be kept as low as possible and PEEP should be limited. The ventilatory rate may be varied to achieve the desired PaCO2 without significantly affecting leakage through the fistula.