Abstract This paper presents scaled physical model studies for petroleum contaminant removal using surfactant solutions. Scaling criteria are selected in such a way that a realistic representation of the decontamination process is possible. The proposed scaling approach meets all important requirements of this decontamination process which uses surfactant solutions. This approach scales more accurately where aquifer geometry, viscous forces, and the ratio of the gravitational forces to viscous forces are important. Experimental runs were conducted to investigate the effects of surfactant concentration, contaminant type, and interfacial tension. The optimum surfactant concentration was identified for one of the contaminants used. It was found that the optimum concentration in the field is not necessarily the same as the one which gives rise to the minimum interfacial tension in a spinning drop tensiometer. This aspect of optimum concentration could not be identified in an unscaled physical model. The only factor that was not modelled is heterogeneity. Scaling criteria or proper characterization of a heterogeneous medium is not yet available. However, the scaled physical model, as presented in this paper, provides a realistic approach to modelling a field remediation process enabling one to design a pilot test for further studies. Experimental results are scaled up to field values. Experimental results indicate that the proposed method of decontamination may recover up to 93% of the contaminant originally in place. The clean-up method is effective for a wide range of permeability of the soil. Also, the effect of a vadose zone was studied for a range of surfactant concentrations. Results indicated that the clean-up method has to be planned according to the soils as well as the contaminant.