Our objective was to use a simple screening model to predict the relative leaching of herbicides in Alberta soils to allow producers the option of choosing herbicides with lower leaching potential. Physical properties for each herbicide were obtained from the literature and the Laskowski model was used to calculate the leaching potential (LP) of the herbicides. Relative LP rankings (LPR) were then created by ranking herbicide LP values on a 1 to 9 scale (1 = no leaching; 9 = high leaching). The leaching rates of nine herbicides (2,4-D, dicamba, MCPA, diclofop, quinclorac, bromoxynil, fenoxaprop, triallate and trifluralin) were then determined on soils from the five major soil zones of Alberta (Brown, Dark Brown, Grey, Dark Grey and Black) using small packed soil columns. Eluate fractions were analyzed using a MSD-GC method. Although there were differences related to soil organic matter content, the relative rates of leaching among the nine herbicides were generally quite consistent. Dicamba, 2,4-D, MCPA and quinclorac leached most readily, followed by bromoxynil, and then diclofop, fenoxaprop and triallate, and finally trifluralin, which did not leach. These soil column results and previous field results validated the LPR for most of the nine herbicides. The LPR did appear to underestimate the leaching of MCPA, bromoxynil and quinclorac. LPR values (1 to 9 scale) are a convenient way to convey herbicide leaching information to producers and could easily be included in herbicide guides along with certain provisos.