Noise pollution is one of the four major pollutions in the world. In order to implement adequate strategies for noise control, assessment of traffic-generated noise is essential in city planning and management. The aim of this study was to determine whether space syntax could improve the predictive power of noise simulation. This paper reports a record linkage study which combined a documentary method with space syntax analysis. It analyses data about traffic flow as well as field-measured and computer-simulated traffic noise in two Bulgarian agglomerations. Our findings suggest that space syntax might have a potential in predicting traffic noise exposure by improving models for noise simulations using specialised software or actual traffic counts. The scientific attention might need to be directed towards space syntax in order to study its further application in current models and algorithms for noise prediction.