OBJECTIVE To improve the accuracy of ventilatory measurements in ventilated newborns by means of a numerical correction when a deadspace free differential measuring method using two pneumotachographs (PNTs) is applied and to investigate the clinical usefulness of this correction procedure. METHODS In vitro study and prospective animal study. METHODS Research laboratory of the Clinic of Neonatology and the Animal Research Laboratory, Charité Hospital Berlin. METHODS Ten newborn piglets, weighing 610-1340 g (median, 930 g), age <12 hrs. METHODS The accuracy of both the deadspace free method and the endotracheal flow measurements (conventional method) was investigated using mechanical lung models. A correction procedure for the deadspace free method was developed considering signal delay time and tube compliance between both PNTs. This method was applied to the piglets measured during partial liquid ventilation (PLV). Measurements were done before and after lung lavage and during 30 and 120 mins of PLV (30 mL/kg body weight perfluorocarbon). RESULTS In vitro measurements showed volume differences between both methods of 8%, 12%, 16%, and 17%, respectively, depending on the distance between the PNTs of 10, 60, 120, and 180 cm. After applying the correction algorithm, the differences decreased to 3%, 0%, -2%, and -8%, respectively. The piglets were measured with 120-cm tube length between the PNTs. The correction algorithm reduced the measured tidal volume before lavage by 7%, after lavage by 14%, 30-min PLV by 12%, and 120-min PLV by 10%, corresponding to the changes in respiratory compliance of 1.2, 0.6, 1.0, and 1.1 mL/cm H2O. CONCLUSIONS The deadspace free method can be advantageously used for continuous measurements in newborns despite much higher technical expense. The correcting procedure improved the accuracy of the volume measurement remarkably, especially for lower respiratory compliance.