The main goal of this study is to compute accurate trajectory and dispersion simulations via better wind analyses provided through the WRF-3DVAR assimilation system of radial winds acquired from a single Doppler radar. Over 1,200 hundred, 12-h HYSPLIT (The Hybrid Single Particle Lagrangian Integrated Trajectory) trajectory and dispersion simulations were performed using various meteorological model gridded outputs as inputs for the Lagrangian model. Five simulation experiments for each day at 00, 06, 12, and 18 UTC were conducted for two different time periods: January and June of 2009: EXNARR, EXRUC13, EXRUC20, EXWRFD1, EXWRFD1D2. The WRF model with two domains was initialized using RUC model analyses and hourly forecasts. The outer domain with 4 km horizontal spacing covers the northern Gulf of Mexico and south central U.S while a nested domain with 1 km resolution was centered at the location of the National Weather Service WSR-88D station (Slidell, LA – KLIX). The three-dimensional variational data assimilation system of WRF was implemented after the radial velocity data were preprocessed by an automatic de-aliasing and noise-removal algorithm. The WRF model with its 3DVAR was run for 29 cases. A detailed characteristics of the resulted trajectory and dispersion simulations for each experiments were documented. As for the trajectory evaluations, absolute horizontal and vertical deviations and relative horizontal deviations were computed. Correlation coefficients, figure of merit in space (FMS), Global analyses (FA2s and FA5s), and Kolmogorov-Smirnov parameter (KSP) were calculated for the concentration simulations between the experiments. In general, correlations of dispersions between control runs and experiments were higher for January simulations while correlations between the control runs and radar-assimilated experiments were the highest. Thus, assimilation of radar radial wind observations had a positive impact on dispersion forecast. Increased in model grid resolution resulted in higher (better) FMS between the experiments and control runs for the month of January as well. June-simulations, on the other hand, did not indicate any improvements in FMS significantly when/if model grid resolutions were increased. Overall, the spread in trajectories between experiments and the control runs varied from 40 km to 60 km within a 12-h simulation window.