Electro-oculography is used as electrodiagnostic test in ophthalmology. The results show considerable variability. For this study a multivariate experiment was designed to help identifying independent variables contributing to the light peak to dark trough ratio and to test which response parameter carries the least variance. No sex difference was found for the peak to trough ratio. Trough and peak potentials are highly correlated. The standard error of the peak to trough ratio is 0.17 or about 10% of the peak to trough ratio. Up to 71 percent of the variance can be accounted for when one light and dark phase of 12.5 minutes each is used. The empirical model is less precise in its predictive value than a formal model, which accounts for more than 90 percent of the variance.