A series of column experiments on the cementation of cadmium ions by zinc powder were conducted as a function of amount of zinc dosage, the initial concentration of cadmium ion, flow rates, depth of bed, and the addition of surfactant to obtain more applicable design data. Cementation of cadmium by the zinc column test was shown to be a feasible treatment process to achieve a high degree of cadmium removal within a fairly reasonable contact time. The efficiency of utilization of a zinc powder bed improved with increasing empty bed contact time (EBCT), but the cementation capacity of zinc powder usually increased with decreasing EBCT value. The breakthrough curve was extended and the effective cementation capacity of the zinc cementation column was raised by adding small quantities of sodium dodcyl sulfonate (SDS) surfactant, probably due to the morphological change involving the cemented cadmium deposits on the zinc surface. Three different adsorption models, the bed-depth-service-time (BDST) model, the Wolborska model, and the Clark model are used to simulate the experimental data. The cementation behavior of cadmium ions in a zinc powder column can be well described by the BDST model, but the fitted rate constant of cementation (k(ads)) decreased with increasing EBCT, thus, it is necessary to modify the BDST model to further estimate the exhaust time of the zinc powder column under various conditions. The fitted results of the other two models indicate higher deviations to the experimental results possibly because the cementation rate was mainly determined by the surface reaction step.