The purposes of this study were to compare pulmonary artery, oral, axillary, and two tympanic temperatures and test agreement between measures of pulmonary artery and tympanic temperatures in a convenience sample of 32 Intensive Care Unit patients. The correlation coefficients for pulmonary artery and oral, axillary, and tympanic temperatures were moderate to strong and significant at the p <.01 level. However, intraclass correlation coefficients estimated the intermethod reliability and were below the .75 criterion of acceptability. Thus, one method of temperature measurement could not be interchanged for another. ©1995 John Wiley & Sons, Inc.