The light brightness test is an investigation in which crossed polarizing lenses placed in front of each eye are rotated with respect to one another to ellicit any disparity in perception of a diffusely illuminated screen. A number of different strategies can be employed to achieve the aim of determining the degree of disparity which is perceived as equal by the patient. A cohort of 45 control subjects was assessed in order to determine the ranges of brightness disparity accepted as equal for six methods of assessment. The method which gave the optimal confidence limit was to preset one pair of lenses at 0 degrees (maximum transmission) and the other as a fixed reference at 45 degrees. Rotation of the lens set at 0 degrees was carried out until the brightness of an X-ray box is perceived as equal. The strategy was then reversed. A disparity in one or both recordings greater than 76.5% is outside 99% confidence limits for the whole population studied. One of 14 patients with ocular hypertension gave a consistently 'positive' results. For the optimal method, over 60% of patients with glaucoma fell outside 95% condidence limits for normal controls indicating that a disparity in brightness perception between eyes is a common feature in glaucoma, in which the disease process usually affects one eye more than the other. For the patients with chronic glaucoma there was a positive correlation between the difference in brightness perception and the difference in central visual field score, indicating that brightness perception may be subserved by the central 30 degrees of the retina. When combined with visual acuity assessment this easily performed test warrants evaluation as a potential screening test for chronic glaucoma.