A new method for measuring and correcting air leaks during lung-function testing in infants has been validated in vitro and in vivo by using a flow-through system that measured the inflow and outflow of a face mask. An adjustable leak was quantified by using suction flow to validate the accuracy of leak measurements. To validate the leak correction, the volume of a pump was measured with different air leaks (0-30%). The method developed was tested in 67 infants breathing spontaneously. There was good agreement between measured and simulated leaks (r = 0.998, P < 0.001; 95% limits of agreement were -0.3 and 0.1%, respectively). The volume was generally underestimated because of leaks, and the volume error was up to 94% compared with the maximum error of 5% after leak correction. With continuous leak measurements in vivo, there were <4% actual leaks (median 2.6%), and we did not observe any leaks in >7% of cases. The leak correction improved the accuracy of ventilatory measurements. The monitoring of leaks is helpful for airtight placement of the face mask and for prevention of serious measurement errors caused by leaks.