The selectivity sequence and removal of coexistent heavy metals (namely As, Cd, Cr, Cu, Ni and Zn) in synthetic urban stormwater runoff were investigated by adsorption onto a porous iron sorbent (namely P1) and its mixtures with zeolite and crystal gravel, respectively (namely P2, P3, and P4). A batch method was employed to simulate the sorption processes. The geochemical model PHREEQC was used to calculate the metals' species and saturation data for elucidating the sorption data. The equilibrium data demonstrated a good fit with the Freundlich model and showed affinity in the orders: Cd > Zn > Ni > Cu > As > Cr (sorbents P1, P3 and P4) and Cd > Zn > Ni > As > Cu > Cr (sorbent P2). In addition to this, Calculated Distribution Coefficient (K-d) values were used to compare the overall heavy metal removal efficiencies of the sorbents, which, in decreasing order. was found to be P4 > P1 > P2 > P3. In comparing these four commercial sorbents, sorbent P4 represents a promising material for treatment of urban stormwater runoff containing mixed heavy metals. (C) 2009 Elsevier B.V. All rights reserved.