Image analysis (IA) techniques are increasingly being used in porous media experiments to measure system properties such as concentration and water content. The values of system properties estimated using IA techniques can be influenced by various types of experimental errors. These errors are generally quantified by using global mass balance calculations or by comparing the dispersion coefficient value obtained from the IA data against an accepted value. We used a theoretical test problem to show that both of these error quantification methods have severe limitations. Hence, we developed an alternative statistics-based method for quantifying IA errors. The applicability of the new method was verified using the theoretical test problem. In addition to quantifying errors, the method can also be used as a design tool for selecting optimal concentration ranges for conducting contaminant transport experiments with minimal errors. We conducted a dense-tracer transport experiment to demonstrate the use of the proposed error analysis method.