The objective of this study was to compare the conventional method of measuring respiratory mechanics, which requires the passage of an esophageal tube, with the occlusion technique, which is less invasive. Thirty-nine preterm infants who received mechanical ventilation on the first day were studied before discharge (mean +/- SD: postnatal age 67 +/- 23 days; weight 1790 +/- 300 gm), and 27 of them again at 1 year (weight 8.1 +/- 1.4 kg). Flows were measured through a nosepiece by pneumotachometry, tidal volume by integration of flow, esophageal pressure through a water-filled tube, and airway pressure directly at the nasal piece. Airway occlusion was performed at the end of inspiration, and the following relaxed exhalation was analyzed to give compliance (Crs) and resistance (Rrs) of the respiratory system. These values were compared with dynamic lung compliance (Cdyn) and expiratory resistance (Re) of the previous unoccluded breath. In the younger infants, dynamic and static measurements did not differ significantly and were well correlated (Cdyn/Crs, r = 0.91; Re/Rrs, r = 0.95). In the older infants, Crs was 80% of Cdyn (p less than 0.001), and Rrs was 24% higher than Re (p less than 0.001). The measurements were well correlated (Cdyn/Crs, r = 0.94; Re/Rrs, r = 0.91). The regression line Cdyn versus Crs had a slope (0.77) significantly less than 1; the regression Re versus Rrs had an intercept (13.8) significantly greater than zero. The lower Crs and higher Rrs values can be expected because the static determinations include the chest wall. In the more immature infants, the very compliant chest wall, in combination with an underestimation of Cdyn because of the higher breathing frequency of these infants, may obscure this difference. We conclude that the occlusion technique gives accurate and reproducible results, is easily applied, does not need the passage of an esophageal tube, and is well tolerated by the infants.