To test the hypotheses that (1) CYP2D6 genotype is associated with pharmacokinetics of ophthalmic timolol and (2) variation in genotypes of ADRB1 (beta(1)-adrenoceptor) and GNAS1 (alpha-subunit of G-protein) modulate heart rate (HR), and systolic (SAP) and diastolic (DAP) arterial pressure responses to timolol.Nineteen glaucoma patients and eighteen healthy volunteers were treated with 0.5% aqueous and 0.1% hydrogel formulations of ophthalmic timolol using a randomised cross-over design. The participants conducted head-up tilt and maximum exercise test at four visits. Plasma concentration of timolol was measured twice for glaucoma patients and ten times for healthy volunteers on each visit. Also, the genotypes for CYP2D6, ADRB1 and GNAS1 were determined.Among healthy volunteers using aqueous timolol, poor metabolisers (PMs, n=2) of CYP2D6 had higher maximum plasma concentrations (C(max), values 2.63 and 2.94 ng/ml), longer elimination half-lives ( T(1/2), 5.49 and 6.75 h), and higher area-under-curve (AUC, 19.54 and 23.25 ng.h/ml) than intermediate [IMs, n=6, mean+/-SD 1.73+/-0.59 ng/ml (not significant), 3.30+/-0.48 h, 11.32+/-3.72 ng.h/ml], extensive (EMs, n=8, 1.60+/-0.72 ng/ml, 3.24+/-1.24 h, 8.52+/-6.12 ng.h/ml) and ultra-rapid (UMs, n=2, values 1.23 and 1.67 ng/ml, 2.22 and 2.52 h, 6.16 and 6.94 ng.h/ml) metabolisers. The IMs, EMs and UMs did not differ from each other for any of the kinetic variables. Also, the elevation of HR from rest to maximum level tended to differ between PMs and IMs, and between PMs and UMs. The pharmacokinetics and pharmacodynamics between the CYP2D6 groups did not differ with statistical significance when hydrogel timolol was used. Upon head-up tilt, the Ser49 homozygotes (n=26) had higher SAP (P=0.03) and DAP (P<0.01) than the Gly carriers (n=11). The change in DAP from rest to maximum during exercise was lower (P<0.01) in subjects with CC alleles of GNAS1 (n=13) than those with at least one T allele (n=24).The CYP2D6 poor metabolisers may be more prone to systemic adverse events with aqueous timolol than extensive metabolisers. Since CYP2D6 genotyping is not routine clinical practice, using 0.1% timolol hydrogel instead of 0.5% aqueous preparation will increase patient safety.