In recent years, operational lidar systems have been developed for ground-based and airborne platforms. These active remote sensing systems are capable of measuring key atmospheric variables with high accuracy and resolution. Consequently, lidar systems have been applied during several field campaigns for process studies. It has also been proposed that water vapor lidar data should have a positive impact on short-range quantitative precipitation forecast (QPF) if these data were assimilated in mesoscale models. However, corresponding data assimilation studies are lacking to date. The International Water Vapor Project (IHOP_2002) has been designed for investigating this statement. This study addresses this issue and introduces the first 4-dimensional variational assimilation of a water vapor differential lidar (DIAL) data in a mesoscale model. A meteorologically complex case from IHOP_2002 was selected. During this day on May 24, 2002, data of the NASA LASE instrument where available in the upstream area of a region where later strong initiation of convection took place. The data were assimilated in the MM5 model by developing a suitable forward observation operator. For the cost function a code was prepared which constructs the corresponding error covariance matrix of the LASE instrument for each profile. We demonstrate a huge impact of the LASE data on the initial water vapor field. This impact persists up to 12 h in the model forecast. Even more striking is a significant improvement of the spatial/temporal prediction of precipitation using the initial water vapor field where the LASE data were assimilated.