Adsorptive bubble-separation processes are used to remove adsorptive contaminants from wastewater. A mathematical model was developed for the adsorptive bubble separation processes in a countercurrent column. The model included an assumption that linear adsorption of the solute occurred on the gas bubbles. Predictions indicated that aqueous column height has less effect on the solute removal efficiency than water flow rates, gas flow rates or bubble size. Increased water flow rates reduce removal efficiency; greater gas flow rates increase removal efficiency. The maximum removal efficiency is determined by adsorption characteristics of the solute; the removal rate is controlled by mass-transfer rate. However, influent concentrations of the solute in the water phase do not affect removal efficiency. These predictions were tested by means of solvent sublation experiments conducted in a laboratory setting. The data were collected from the sublation of Triton-X100 in a water-ethyl acetate-nitrogen system. Linear regression was used to compare predictions and experimental data with respect to column height. Five column heights ranged from 30 to 100 cm were compared at three water flow rates, 3.0, 5.5 and 13.0 ml/min. The calculated column heights agreed the observed column heights well. As the model predicted, the effect of influent concentraion on the removal efficiency was not significant. The surface excess of Triton-X100 (0.75 E-10 to 7.5 E-10 g mole/cm2) estimated from the adsorption constant in the model agrees with other published values (2.6 E-10 to 2.8 E-10 g mole/cm2). The mass-transfer coefficients of Triton-X100 in water (kL, which ranged from 0.50 to 0.60 cm/min) determined from the model and experimental data at 3.0, 5.5, 8.0 and 13.0 ml/min water flows (Qw) are proportional to the 0.124 power of the water flow rates. kL = 0.0441 Qw0.124 This empirical exponent correlation between mass-transfer coefficients and water flow rates for the adsorption processes is similar to the correlation determined from previous absorption models.