Subject acceptance of a single dose of timolol and levobunolol was compared in a three-center, double-masked, randomized, crossover study in healthy normal subjects. We tested 115 subjects who received the medications OU, in a random order, separated by a washout period of three to seven days. At the end of the administration, the subjects were asked to report any symptoms and were monitored by ophthalmologic examination. Those receiving levobunolol reported a two- to threefold higher frequency of symptoms than did those receiving timolol (P = .0002, by Mainland-Gart chi-square test). Thirty-five subjects (30.4%) treated with levobunolol complained of burning versus 14 (12.2%) receiving timolol (P = .0021). Similar results were obtained in regard to stinging (31 versus 10 subjects; P = .0011). The duration of these symptoms did not differ between the two groups (P > .05). Timolol induced considerably less burning and stinging on day 1 compared with levobunolol, using a visual analog scale over time (P = .0004). Seventy-nine percent of those stating a preference chose timolol (P < 3 x 10(-9).