During the course of negative provocative test for closed-angle glaucoma using pilocarpine and phenylephrine 60% of eyes develop significant reductions in outflow facility at some stage during the test. It is shown that these reductions can be explained by postulating the presence of partial-angle closure since: (1) A random sample (6) of 53 eyes showing an abnormal response subsequently had a peripheral iridectomy. On reprovoking they then behaved as normal eyes with a uniform increase in outflow. (2) Fifty-eight eyes that had a peripheral iridectomy for closed-angle glaucoma (spontaneous or induced) responded to provocative testing as do normal eyes.