A new microwave profiling radiometer manufactured by Radiometrics Corporation was deployed at Lamont, Oklahoma from 15 February to 8 August 2000. The microwave radiometer profiler (MWRP) provides continuous (~10-min temporal resolution) measurements of atmospheric temperature and humidity up to 10 km in clear and cloudy conditions, and a liquid water content (LWC) profile of one cloud layer. Diurnal variations in atmospheric water vapor, liquid water, and temperature profiles along with precipitable water vapor (PWV) and liquid water path (LWP) are studied by analyzing this high-temporal-resolution MWRP dataset. We compare MWRP-derived data from spring and summer of 2000 with 3-hourly ARM radiosonde data from Lamont from 1994 to 2000. Water vapor mixing ratio (MR) in the upper troposphere (above ~ 6 km) in both spring and summer is significantly higher in the early morning (00-08 local solar time LST) than during the day (08-18 LST). In the lower troposphere (below 3 km), water vapor MR in both seasons tends to be lower in the morning than in the afternoon and the night, and reaches a minimum around 08 LST and a maximum around 18 LST. This minimum is propagated upward from the near surface, while the late-afternoon maximum propagates downward to the near surface. In the middle troposphere (3~6 km) water vapor MR shows complicated diurnal variation, with different patterns in spring and summer. Relative humidity (RH) shows the same diurnal variations as MR in the middle and upper troposphere, but is out of phase with MR in the lower troposphere. The latter is attributed to strong temperature diurnal variation in the lower troposphere. PWV peaks around 17 LST in both spring and summer. Diurnal variations in the LWC and LWP are small and noisy. MWRP-derived diurnal variations in MR, RH and PWV are consistent with those from 3-hourly radiosonde data during the 1994-2000 period except in the middle troposphere in summer. Vertical profiles of water vapor and temperature anomalies at 05, 14, 17 and 23 LST were calculated from the radiosonde data from March to August 2000 and compared with the MWRP data. These two datasets show good agreements for the whole profiles in spring and below 2 km at 05, 14 and 17 LST in summer. Radiosonde-calculated RH anomalies above 2 km in summer are much larger than those from MWRP data. This is likely due to missing MWRP data during rain conditions. Possible physical processes responsible for observed diurnal variations will be discussed.