From hydrocarbon reservoirs, brine is produced as a waste material, which may be injected into the ground or discharged at the surface. When the wastewater is injected into the ground, it may be mixed with fresh-water sources by several processes. Groundwater contamination from leakage, spills, or the injection of hazardous or toxic materials is widely regarded as one of the leading environmental problems. This study presents the use of genetic algorithms (GAs) as a viable means of estimating the transport parameters such as dispersivity, retardation factor, and diffusion coefficient of water-saturated porous media. The unknown transport parameters of advective-dispersive contaminant equations for homogeneous, linear, radial, and fractured systems are predicted by the use of GAs coupled with the experimental data. The parameter estimation study is considered as a constrained optimisation problem by minimising the total error between the calculated and the measured effluent concentrations satisfying state equations, boundary conditions, and limits on parameters. The optimisation formulations combine advective-dispersive pollutant transport simulation with GAs optimisation. The calculated concentrations were obtained from analytical solutions of these models. The results of the simulation runs were calibrated with those obtained from the experimental tests. A satisfactory agreement between estimated and experimental results was achieved. Since all the unknown parameters in the models are found to be in a reasonable agreement with respect to the experimental conditions, GAs can be said to be an effective optimisation method for estimating the transport parameters in groundwater contamination problems.