The authors compare the results of computer simulated flow fields around a complex building (B170) at Lawrence Livermore National Laboratory (LLNL) with field measurements. This is the first stage of a large effort to assess the ability of computational fluid dynamics (CFD) models to predict atmospheric dispersion scenarios around building complexes. At this stage, the focus is on accurate simulation of the velocity field. Two types of simulations were performed: predictive and post-experiment. The purpose of the predictive runs was primarily to provide initial guidance for the planning of the experiment. By developing an approximate understanding of the major features of the flow field, they were able to more effectively deploy the sensors. The post-experiment runs were performed for several reason: (1) the largest amount of experimental data was available for slightly different wind directions than the directions used in the initial calculations. The predictive runs simulated three wind directions: 200, 225, and 250 degrees measured from true north. Although, the winds did blow generally from the southwest (typical summer conditions for this site), the most appropriate data available was for 210, 225, and 240 degrees. (2) They wanted to explore the sensitivity of the predictions to various levels of idealization that are by necessity a part of the modeling process. For example, what level of detail is required to accurately model the effect of the trees? How much architectural detail should be included in the model of the building? (3) They are testing the sensitivity of the results to different turbulence closures.