The usefulness of Nimbus 4 satellite infrared spectrometer (SIRS)-derived temperature and height data for constant-pressure analyses at stratospheric levels is investigated by comparing SIRS data with rawinsonde observations and objective analyses of those data. Results from the various methods of comparison are difficult to interpret since systematic and random errors of observatfons at stratospheric altitudes do not permit the observed data to be used as an unquestioned standard. In addition, conclusions must be qualified by the fact that the SIRS information derived to date has depended in part on analyses of rawinsonde data. The following conclusions were reached from the various comparison studies: (1) SIRS data are useful for constantpressure analyses at stratospheric levels, (2) mean differences between analyzed rawinsonde temperatures and SIRS derivations are'generally less than 3"C, (3) differences are greater during stratospheric warmings, but SIRS data generally indicate the proper trend of the temperature changes, thus adding information about the temperature of the real atmosphere to an analysis, and (4) stratospheric SIRS data after Nov. 5, 1971, can be used with more confidence than those derived before that date.