Transport of solute through soils is an important topic of current interest in groundwater hydrology and environmental geotechnics. Laboratory column test is usually conducted to determine transport parameters that are key to analytical solution of the advection-dispersion equation. Significant uncertainty often exists in the solute transport parameters, as they are obtained empirically through nonlinear curve-fitting of column test data. In the present study, a series of probabilistic sensitivity analysis is carried out to investigate the effects of transport parameter uncertainties on the effluent concentration in a column test. The probabilistic sensitivity analyses are performed using @Risk, a macro running on Excel, a popular spreadsheet program. @Risk allows the user to place a random variable in any spreadsheet cell. The simulation analysis in @Risk is based on Monte Carlo or Latin Hypercube sampling technique. Various issues on the effect of transport parameter uncertainties are investigated. Based on the findings of the probabilistic sensitivity analyses, an improved method for more accurate and consistent determination of the transport parameters from the column test is developed. A column test is conducted and analyzed using the traditional method and the proposed method and a comparison of the two methods is made. The advantages of the proposed method are demonstrated.