Long-term use of drugs that suppress aqueous humor formation, such as timolol and dorzolamide, or that redirect aqueous humor outflow from the trabecular meshwork, such as prostaglandin F 2α analogues, could cause underperfusion of the trabecular meshwork and a secondary decrease in outflow facility. We investigated the mechanism of suppression of aqueous humor formation by timolol in monkey eyes by measuring aqueous humor ascorbate levels. We also determined whether suppression of aqueous humor formation with and without redirection of aqueous humor away from the trabecular meshwork could lead to a subsequent reduction in outflow facility, and whether this reduction was correlated with increased fibronectin levels in anterior chamber aqueous humor. In cynomolgus monkeys, unilateral dose/aqueous humor formation response curves were generated for timolol, dorzolamide, and a combination of timolol+dorzolamide. Aqueous humor formation and/or outflow facility were measured in both eyes after approximately four days, four weeks and seven weeks of twice daily treatment with 3·5 μg timolol+1·0 mg dorzolamide to one eye and 30% DMSO to the other. In some monkeys, 5 μg prostaglandin F 2α-isopropyl ester (PG) was added to timolol+dorzolamide for 4-week treatments. Intraocular pressure and corneal endothelial transfer coefficients ( k a) were also measured at four weeks. Aqueous humor fibronectin levels were determined in four monkeys after approximately 9·5 weeks of timolol+dorzolamide treatment. Aqueous humor formation, intraocular pressure, and aqueous humor ascorbate levels were also determined in rhesus monkeys at baseline and after a single unilateral topical administration of 25 μg timolol. Compared to baseline for the same eye, aqueous humor formation was significantly decreased in treated eyes at all doses of timolol and at 1·8 and 4 mg dorzolamide. Compared to the opposite control eye, aqueous humor formation was lower in treated eyes after 3·5 and 5 μg timolol and after all doses of dorzolamide. Aqueous humor formation after treatment with 3·5 μg timolol+1·0 mg dorzolamide was decreased in treated vs. control eyes, after four days and was suppressed in both treated and control eyes after four weeks of treatment, but not when PG was added. There was no difference in k a values with or without the addition of PG. Intraocular pressure was significantly lower in both treated and control eyes vs. baseline after ∼6·5 weeks treatment with timolol+dorzolamide when taken 2 hr after the last dose and after ∼3·5 weeks treatment with timolol+dorzolamide+PG when measured 6 hr after the last dose. Outflow facility after treatment with timolol+dorzolamide was unchanged after four days, tended to be lower in the treated vs. control eyes after four and seven weeks, and was significantly lower in treated vs. control eyes after four weeks treatment with timolol+dorzolamide+PG (0·352±0·052 vs. 0·515±0·096 μl min −1 mmHg −1, p≤0·02). Both treated vs. control eye aqueous humor fibronectin levels were below the level of detection for our assay (0·01 μg ml −1). The 25 μg timolol dose decreased ipsilateral, but not contralateral intraocular pressure (12·6±1·7 vs. 15·2±0·9; p<0·05) and aqueous humor formation (1·40±0·08 vs. 2·03±0·09 μg ml −1, p≤0·01). There was no difference in anterior chamber ascorbate levels in treated vs. control eyes or compared to their respective baselines. Our findings indicate that timolol affects neither ciliary epithelial transport of ascorbate nor aqueous fibronectin levels. Our data also indicate that decreasing aqueous humor formation over a period of time can lead to reduction in outflow facility, particularly when combined with therapy that redirects aqueous from the trabecular meshwork. Future intraocular pressure-lowering therapies for glaucoma may better be directed at enhancing flow through the trabecular pathway as opposed to decreasing aqueous humor formation or rerouting aqueous humor away from the trabecular meshwork.