Methods for reducing and analyzing the measurements made of splash and spray from heavy trucks using laser transmissometers have been under study for the past decade. Approaches have ranged from simple selection of spray density data with no corrections for wind direction and velocity to elaborate “normalization” approaches. None of these approaches have assisted in predicting what a given aerodynamic configuration will produce under wind conditions that have not actually been investigated. This paper describes an approach to such prediction based on a regression model that was developed from well over 100 observations made on one configuration under a wide variety of different wind conditions. The model is then applied to recent test results to illustrate its utility in making splash and spray evaluations more generalizable to real-world conditions.