Finding the eyes is an important stage of feature extraction in automatic face recognition. Current approaches include standard feature extraction techniques using heuristic methods specifically developed for human eyes. We present a new method for finding eye centres using a gradient decomposed Hough Transform (HT) which embodies the natural concentricity of the eye region in a peak reinforcement scheme to improve accuracy and robustness. This enhances a standard feature extraction technique with an analytic approach, which can be applied to the whole face without priming of estimates of eye position and size. In a database of 54 eyes this new method is shown to be less constrained, more robust and resulted in a three-fold improvement in accuracy over using the standard HT.