The majority of shunt dysfunctions can be easily diagnosed by puncturing the Rickham reservoir with a fine needle and estimating the intracranial pressure. Experience has, however, shown that definite shunt dysfunctions can be missed by this method. These limitations of the technique are better understood by constructing a model which enables us to simulate the pressure and flow situation in a ventricular drainage system. This experimental model has shown that in cases of partial proximal or distal blockage, the flow through the shunt system is significantly diminished but not totally stopped, and the pressure reading may thus be normal. In these cases, the shunt should be revised on the basis of clinical evidence of dysfunction.