We studied reproducibility and variability of dynamic pulmonary compliance (Cdyn) by making measurements with the esophageal balloon at multiple locations within the esophagus, in both spontaneously breathing and mechanically ventilated newborn infants. Reliable measurements could be obtained over a range similar to that reported for measurements with a liquid-filled catheter. In spontaneously breathing infants Cdyn was found to be highly variable. This variability was unrelated to catheter position but was associated with concomitant changes in pulmonary resistance. Probably because of the high variability, the correlation of Cdyn with a measurement of respiratory system compliance (Crs) was rather poor (r = 0.63). Cdyn measured in mechanically ventilated infants was significantly less variable and compared favorably to Crs (r = 0.86), but its accuracy could not be adequately assessed since the comparison of esophageal and airway occlusion pressure was not feasible in all infants. In addition, significant differences in Cdyn were found between spontaneous and ventilated breaths during mechanical ventilation. Further studies in both ventilated and spontaneously breathing infants are needed to assess the variability of Cdyn over extended time periods.