We devised a new calibration procedure [least squares method (LSQ)] for respiratory inductive plethysmography (RIP) and compared it with our previously reported simultaneous equation method (SEQ) of analyzing data in 2 body positions and with the method of Stagg and associates using the analysis of individual breaths in a single body position. The values from RIP were compared with simultaneous spirometry (SP) in 20 normal subjects placed in the standing (STD), supine (SUP), sitting, prone, semi-recumbent, right lateral decubitus, and left lateral decubitus postures. The LSQ gave the most accurate results followed closely by SEQ. In addition, LSQ was compared with the isovolume angle maneuver (ISV) calibration procedure in supine (ISV-SUP) and standing (ISV-STD) postures. Each of the 10 normal subjects breathed at tidal volumes of 250, 750, and 1,250 ml in the SUP and STD postures. Of the values obtained by the LSQ method, 93% were within +/- 10% of SP in SUP and STD positions. Without a change in the posture in which the calibration was made, 83% of values with ISV-SUP and 90% of values with ISV-STD were within +/- 10% of SP. When body position was changed, 65% of the values obtained with ISV-SUP and 38% of the values obtained with ISV-STD were within +/- 10% of SP. With the LSQ, 45% of isovolume angles in SUP and STD position were within 45 +/- 3 degrees; 40% of isovolume angles with ISV-SUP and 60% with ISV-STD were within 45 +/- 3 degrees when body position was changed from position calibrated. In estimating fractional contribution of rib cage and abdominal compartments. LSQ was comparable to ISV in the standing posture but generally gave lesser values for the rib cage contribution in the supine posture than ISV. The optimal calibration procedure for respiratory inductive plethysmography in terms of accuracy and ease of subject performance is the least squares calibration procedure.