An evaluation of continuous intracranial pressure (ICP) monitoring in 42 hydrocephalic infants and children helps to establish the role of this method in comparison with other clinical techniques for estimating the occurrence of progressive hydrocephalus. Transducer monitoring via ventricular catheters in term and premature infants revealed average ICP in the 5--12 mm Hg range. The higher values were recorded from patients with noncommunicating hydrocephalus but did not correlate with the extent of ventriculomagaly. The method has very litte predictive value in estimating the progressive nature of infantile hydrocephalus. In older children ICP monitoring provides a useful adjunct in decisions for shunting after intracranial surgery and in suspected 'normal pressure' hydrocephalus. Monitoring from shunt reservoirs in cases of suspected shunt obstruction, although accurate in most situations, is probably inferior to clinical examination and serial computed tomography considering the expense and the risk of introducing shunt sepsis.