In this study, a clinoptilolite bed column system was used to remove ammonium from municipal landfill leachate. Laboratory-scale column experiments were conducted in upflow fixed-bed and fluidized-bed modes with different ammonium concentrations. Ammonium removal was managed mainly by Ca2+ and K+ cations of clinoptilolite, and higher treatment performances were obtained at lower flow rates. On the other hand, higher effluent volumes and removal rates were produced by lower ammonium concentrations, and increased expansion ratios in the fluidized-bed column reduced the treatment efficiency. The sum of normalized errors (SNE) procedure was applied using five different error functions to model the experimental data. The Clark and Yoon-Nelson kinetic models were applied to column data to predict the breakthrough data, and the results indicated that the Clark model is better for modeling the experimental data. Ammonia stripping was investigated as a pretreatment step to obtain lower ammonium concentrations in the effluent. With a 12-h aeration time at pH 12, the ammonium concentration of leachate was decreased from about 3000 to below 400 mg/L. These results indicate that combining ammonia stripping and an upflow fixed-bed system is feasible in reducing higher ammonium concentrations to satisfactory levels.