Children with cortical visual impairment (CVI) usually have a typical clinical presentation. However, in some cases, it may be useful to have confirmatory evidence based on objective electrophysiologic information. To achieve this, we examined some mathematically derived parameters constructed from 20 channel visual evoked potential (VEP). A group of 30 children diagnosed with CVI by clinical and CT findings was compared to a normal control group of 52 children. Each recorded VEP was mathematically transformed using Hjorth's source derivation, to reduce reference contamination and enhance local features. The area under the response curve, computed for each channel within a fixed time window, was used as a measure of the response activity at that channel. These areas were then used to construct several parameters ("R values") describing ratios of activities between different recording electrode areas. Some of these ratios provided good separation between patient and control groups, especially for children older than 5 years of age; in particular CVI patients were found to have a low occipital-to-parietal activity ratio. This finding, together with the observed age independence of the R values in the normal population, their ease of computation and possible physiological interpretability, suggest that R values could be used as confirmatory diagnostic measures.