This study investigates the benefits of a satellite HYper-spectral Microwave Sensor (HYMS) for the retrieval of atmospheric temperature and humidity profiles, in the context of Numerical Weather Prediction (NWP). In the infrared, hyper-spectral instruments have already improved the accuracy of NWP forecasts. Microwave instruments so far only provide observations for a limited number of carefully selected channels. An information content analysis is conducted here to assess the impact of hyper-spectral microwave measurements on the retrieval of temperature and water vapor profiles under clear-sky conditions. It uses radiative transfer simulations over a large variety of atmospheric situations. It accounts for realistic observation (instrument and radiative transfer) noise and for a priori information assumptions compatible with NWP practices. The estimated retrieval performance of the HYMS instrument is compared to those of the microwave instruments to be deployed on board the future generation of European operational meteorological satellites (MetOp-SG). The results confirm the positive impact of a HYMS instrument on the atmospheric profiling capabilities compared to MetOp-SG. Temperature retrieval uncertainty, compared to a priori information, is reduced by 2 to 10%, depending on the atmospheric height, and improvement rates are much higher than what will be obtained with MetOp-SG. For humidity sounding these improvements can reach 30%, a significant benefit as compared to MetOp-SG results especially below 250 hPa. The results are not very sensitive to the instrument noise, under our assumptions. The main impact provided by the hyper-spectral information originates from the higher resolution in the O2 band around 60 GHz. The results are presented over ocean at nadir but similar conclusions are obtained for other incidence angles and over land.