Simultaneous measurement of esophageal and tracheal pressures during an occluded inspiratory effort was used to assess the accuracy of the esophageal balloon for measuring pleural pressure in dogs. Esophageal balloons were inserted in five mongrel dogs, and an occlusion test was performed with the balloon tip 5, 10, 15, 20, and 25 cm above the esophageal sphincter; at lung volumes of functional residual capacity (FRC) and FRC + 600 ml; and in supine and right- and left-side lying postures. The protocol was repeated in paralyzed animals. This time the occlusion test was performed by injecting air into a plethysmograph to change the body surface pressure, simulating pressure changes produced by respiratory efforts in spontaneously breathing animals. In 47% of the tests in spontaneously breathing dogs, the slope of esophageal vs. tracheal pressure varied greater than 10% from unity. After paralysis the slope did not vary greater than 5% from unity under any circumstance. These data indicate that the poorer performance of the occlusion test in nonparalyzed dogs is due to active tension in the walls of the esophagus and stress induced in the intrathoracic soft tissues by the descent of the diaphragm during a breathing effort.