Soot sensors are used for the diagnosis of Diesel Particulate Filter (DPF) failures that will result in exceedance of Particulate Matter (PM) emission limits defined by increasingly stringent regulations for diesel vehicles. Accumulating sensors, also known as resistive electrode sensors, are considered to be a practical and low-cost approach to estimate soot concentration in diesel engine exhaust, as part of the vehicle On-Board Diagnostics (OBD). This paper presents a physical model describing soot particle deposition mechanisms that was developed to interpret and predict the soot sensor behavior. Initially, the range of parameters that affect sensor output is defined after analyzing vehicle driving cycle measurements. Next, a set of steady-state measurements is conducted and Computational Fluid Dynamics (CFD) simulations are performed to investigate the flow field inside the sensor tip. Data from measurements and simulations feed the developed physical model of the soot deposition on the sensor electrode plate. The investigated deposition mechanisms include thermophoresis, electrophoresis, inertial impaction, and convective diffusion. The model results are validated with several sets of measurements of a diesel engine equipped with DPF based after-treatment system. The developed model is found capable of predicting the behavior of the sensor for a range of exhaust gas conditions, taking into account sensor geometry.