A receptor-source model is used to apportion source contributions for polycyclic aromatic hydrocarbons (PAHs) in street and creek sediments for Sault Ste. Marie, Ontario. The receptor model is shown to overcome the problem of nonlinearity in system modelling. The modelling shows that although the estimated vehicular emissions have the lowest emission rate of PAHs in Sault Ste. Marie, they are a major contributor, along with the coke ovens, to PAH levels in street sediments. The result of source contributions are shown to follow the industrial activities and vehicular traffic densities, which in turn can be used to improve the cost effectiveness of urban runoff control strategies for mitigation of PAH inputs to receiving waterbodies.