Pharyngeal patency is influenced by the surface tension (γ) of the upper airway lining liquid (UAL), of which saliva is a major component. We investigated the influences of saliva production on γ of the UAL, and upper airway re-opening and closing pressures. In 10 supine, male, anaesthetized, tracheostomised, mechanically ventilated New Zealand White rabbits, we measured re-opening and closing of the passive isolated upper airway at baseline and following graded (cumulative) doses of methacholine or atropine. Upper airway liquid volume index (UALVI) was assessed using a standardized suction procedure (secretion weight obtained per second) expressed as the natural logarithm (LnUALVI). The γ of UAL samples were measured using the 'pull-off' force technique. Across all animals, baseline values were: LnUALVI −6.2 (−8.6 to −5.4) median (interquartile range), γ of UAL 58.9 (56.6-59.9) mN m −1 , re-opening 8.6 (6.9-11.1) cmH 2 O, and closing pressures 3.2 (1.8-5.7) cmH 2 O. LnUALVI increased by ∼0.17 per μg kg −1 methacholine and decreased by ∼0.14 per 100 μg kg −1 atropine (both P < 0.03, linear mixed effects modelling). Surface tension was unchanged by methacholine but increased by ∼0.6 mN m −1 per 100 μg kg −1 atropine (P < 0.004). When data were analysed across all animals, both re-opening and closing pressures increased as surface tension increased (by ∼0.4 cmH 2 O mN −1 and by ∼0.7 cmH 2 O mN −1 , respectively; both P < 0.05). We conclude that saliva production influences upper airway mechanical properties partly via alterations in γ of UAL. We speculate that in obstructive sleep apnoea, altered autonomic activity may reduce saliva production and increase surface tension of the upper airway lining liquid, thus increasing the likelihood of upper airway obstruction.