Diaphragmatic force-length behavior was assessed in anesthetized dogs using two techniques. One employed measurements in a diaphragm strip with intact nerve and blood supply; the second related transdiaphragmatic pressure (Pdi) to direct estimates of diaphragmatic tension from strain gauge archesduring bilateral, tetanic, supramaximal phrenic nerve stimulation. In strippreparations the diaphragm exerted active force at lengths as short as 40% of resting in situ length (Lo), and maximal force was registered at a length averaging 25% greater than Lo. This suggested a broader effective length range for the diaphragm as compared to other skeletal muscles. In the intact preparation both Pdi and directly measured diaphragmatic tension were inversely related to lung volume and when they were used to calculate the diaphragm's effective radius of curvature (r), r was found to change little or decrease at large lung volumes rather than increase as theory wouldpredict. These findings suggest that length tension characteristics outweighgeometric considerations in explaining the diaphragm's function in normal dogs and probably normal men.