In a retrospective study we reviewed the records of 788 subjects who had been corticosteroid tested with 0.1% dexamethasone four times daily to one eye for six weeks. All subjects had normal kinetic visual fields and optic nerve heads in both eyes at the time of testing and were followed up for a minimum of five years. Some subjects had normal baseline intraocular pressures whereas others were considered to have ocular hypertension. Of 276 individuals who were high corticosteroid responders (intraocular pressure greater than 31 mm Hg during dexamethasone administration), 36 (13.0%) developed glaucomatous visual field loss during the followup period. Only nine of 261 individuals (3.4%) who were intermediate responders (intraocular pressure 20 to 31 mm Hg during dexamethasone administration) and none of 251 individuals who were low responders (intraocular pressure less than 20 mm Hg during dexamethasone administration) developed glaucomatous visual field loss. However, the ability of the intraocular pressure response to dexamethasone to predict the development of glaucomatous visual field loss was not as good as the predictive power of a multivariate model that included patient age, race, baseline intraocular pressure, baseline outflow facility, baseline cup/disk ratio, and systemic hypertension.