A retrospective follow-up study of 134 patients (84 women and 50 men) tested for their responsiveness to topically administered corticosteroids was carried out. Of the 134 patients tested, 34 (25.4%) had a high response (IOP rise greater than or equal to 16 mmHg), 66 (49.3%) an intermediate (IOP rise 6-15 mmHg) and 34 (25.4%) a low response (IOP rise less than or equal to 5 mmHg) to topically administered 0.1% dexamethasone eye drops. Follow-up time varied from 5 to 15 years (mean 7.1 years). During the follow-up period 29 (21.6%) patients developed glaucoma. Of these, 9 (31.0%) were high responders and 7 (24.0%) had a low response to corticosteroids. Log-linear analysis showed no correlation between corticosteroid response and development of glaucoma in the study eye (SD = 0.232, DF = 2, P = 0.890), the control eye (SD = 1.234, DF = 2, P = 0.540) or both eyes (SD = 0.618, DF = 2, P = 0.734). At the final examination 75 patients (56%) had an IOP of greater than or equal to 22 mmHg. Of these ocular hypertensive patients 28 (37.3%) were high responders, and 15 (20%) had a low response to corticosteroids. Of the whole material of 134 patients 44 had a family history of glaucoma. Eleven (25.1%) of these were high responders and 14 (31.8%) low responders. Of the 29 patients who developed glaucoma 8 had a family history of glaucoma. Of these 3 were high responders and 3 low responders. The results indicate that the dexaprovocative test yielded relatively limited value in the screening of patients for glaucoma.