Polzer et al.’s method combined with Bromley’s method for estimating activity coefficients and a Langmuir isotherm for cesium in a simple simulated waste solution containing 5.1 M NaNO3 and 0.6 M NaOH was used to estimate distribution coefficients for cesium in a complex simulated waste solution characteristic of the radioactive tank wastes at Hanford and other U.S. Department of Energy sites. The ion exchange material was a hydrous sodium crystalline silicotitanate, labeled TAM-5, which is being developed by Texas A&M University, Sandia National Laboratories, and UOP Associates. Cesium distribution coefficients collected by Pray et al. on a NCAW simulated waste solution were predicted with deviations of less than 25% for solutions containing 1 M, 3 M, and 5 M Na+ and Na:Cs ratios of 10(3)-10(8). The deviations were less than 5% for the solutions with 1 M Na+. Cesium distribution coefficients were also predicted and compared with values measured by Egan et al. for TAM-5 and for a storage tank supernate and a "newly generated" waste solution. Excellent results were obtained for the "newly generated" waste simulated solution, which did not contain potassium or rubidium. The predictions for the other simulated waste solution were significantly greater than the measured values, because of the presence of large concentrations of potassium or rubidium. The effect of competitive ion exchange between Os, Rb, and K was not included in the theory. However, the effect of competitive exchange of Cs, Rb, and K appears to be greater for the Oak Ridge simulated waste solution than for the NCAW waste.