Previous research of air quality in the vicinity of roadways has used either field measurements, where results are site specific, or modeling, which can be computationally expensive or too simplified and misleading. The presented research consists of a systematical laboratory and numerical investigation of the influence of different sound barrier (SB) configurations on roadway dispersion. The laboratory experiments are being conducted at the Laboratory for Environmental Flow Modeling at the University of California, Riverside which has a custom made water channel with the capability of performing Particle Image Velocimetry and Planar Laser Induced Fluorescence. The water channel is also being used for visualizations by releasing buoyant dye, a mixture of dye and alcohol. The presented measurements have been made with a neutrally buoyant dye. The laboratory experiments are accompanied with numerical modeling using Quick Urban and Industrial Complex, which consists of a wind and a dispersion module. The wind module solves the Reynolds Averaged Navier-Stokes equations and the dispersion module deploys Lagrangian particle tracking. It was found that in both the water channel and QUIC simulations the presence of SBs increases the vertical spread of the plume relative to that of no SBs. The presence of a SB on the downwind side of the roadway produces a recirculation cavity on the lee side of the SB, and due to this recirculation there is a large decrease in ground level concentrations near the roadway, up to 2 SB heights downwind of the SB. The presence of SBs will decrease the downwind concentrations due to the increased vertical mixing. The effects of roadway turbulence, buoyant release and the effects of ground heating on the dispersion of pollutants will be discussed.