We used data from three Microwave radiometers (MWRs) at 23.8 and 31.4 GHz, radiosonde, and Global Positioning System (GPS) to analyze the performance of five clear-sky atmospheric absorption models. Data were collected during the Cloud InterComparison experiment at the Department of Energy’s Atmospheric Radiation Measurement Program’s (ARM) site in Oklahoma in 2003. We compared brightness temperatures (TBs) measured by the MWRs to the TBs computed from the radiosondes and the absorption models. The radiometer calibration was performed by using two tipping calibration procedures: the ARM calibration algorithm and the tip cal method of the Environmental Technology Laboratory (ETL). Clear skies were determined by using lidar measurements. We found a bias in general less than 0.5 K when the newer models were considered. Precipitable water vapor (PWV) derived from a GPS receiver was also compared to the PWV computed from the radiometers using the retrieval coefficients derived from the models.