Surgical treatment of infantile hydrocephalus by shunt implantation may result in suboptimal intracranial pressure. Major neurological impairments and death are usually prevented by shunt treatment, but minor sequelae may persist or develop. The introduction of adjustable shunts has improved the possibilities of optimizing shunt function and minimizing the risk of such impairments. However, it is still impossible to determine the intracranial pressure without invasive measurements. Clinical findings and procedures such as computed tomography (CT) are not always enough to allow a conclusion as to whether a child's signs and symptoms are the result of suboptimal intracranial pressure (shunt dysfunction) or are of another etiology. With the aim of reducing the number of invasive pressure measurements and CT scans, we investigated the effect of increased intracranial pressure on the visual evoked response (VER). Binocular light flash stimuli of supramaximal intensity were used and VER recordings were performed from Oz and Cz. The VER results from a group of 31 infants and children with hydrocephalus and 2 children with pseudotumor cerebri were compared with responses from a control group of 35 healthy children. The results show that a subpotential, P' (P-prime), usually just preceding P1 (P100), had an increased latency ( > 96 ms) in all hydrocephalic children before surgery. The P' latency in this group was usually even above 110 ms. The latencies of other VER potentials were also increased but not as consistently as P'. After surgical intervention the VER latencies decreased and usually normalized. The P' latency in four children in the control group was just above the borderline latency, but was less than 110 ms.(ABSTRACT TRUNCATED AT 250 WORDS)