Significant progress has been made over the past decade in the development of remote-sensing instruments to profile wind and temperature. However, the current technology of profiling water vapor remotely is still far from perfect. Although some promising optical research systems, such as the Raman lidar, can provide high vertical resolution profiles of water vapor, it may be years before they are generally available. Currently, there are several systems that can measure the vertically integrated water vapor (i.e., precipitable water) with a high degree of accuracy. In this paper we use a simple method to assimilate precipitable water measurements (possibly from a network of dual-channel ground-based microwave radiometers or a satellite-based system) into a mesoscale model. The basic idea is to relax the predicted precipitable water toward the observed value, while retaining the vertical structure of the model humidity field. We test this method with the special 3-h soundings available from the Severe Environmental Storms and Mesoscale Experiment. The results show that the assimilation of precipitable water into a mesoscale model recovers the vertical structure of water vapor with an accuracy much higher than that from statistical retrieval based on climatology. The improved analysis due to assimilation also leads to improved short-range precipitation forecasts.