The nuclear research center at Mol, Belgium, is underlain by clean quartz sand to a depth of about 30 meters. An accident at one of the nuclear facilities, which could contaminate the ground water, would entail serious consequences, since the movement of radioactive contaminants through the ground would be retarded or reduced only slightly by adsorption. Test drilling showed that the water table aquifer is composed of two sand formations, the upper Mol sand being somewhat coarser grained and therefore probably more permeable than the underlying Casterle sand. The lower part of the Mol sand contains a number of thin clay layers which make a partial hydrologic barrier between the two formations, dividing the water table aquifer into two parts. A test well was installed, screened through the full thickness of both sands, and test pumped. Drawdowns were measured in observation wells screened both in the Mol sand and in the Casterle sand. A tracer was added to four of the observation wells in turn, and the arrival of the tracer was measured in the pumped well. The average travel times in the Mol and Casterle sand were used to determine the amount of water being produced from each by the pumping, and these data, with the observed drawdowns, were used to calculate the transmissibility, the permeability, and the storage coefficient of each of the two formations. Using these parameters, the designs of several well systems for pumping contaminated ground water out of the aquifers are discussed. Pumping out, or scavenging, the contaminated ground water and decontaminating it appears to be a practical method of resolving some of the problems that would be raised by a nuclear accident in the Mol area.