In the literature, it remains uncertain whether methylxanthines inhibit hypoxic pulmonary vasoconstriction. We examined the effects of theophylline and of S 9795, a new methylxanthine derivative, on multipoint mean pulmonary arterial pressure (Ppa)/cardiac index (Q) relationships in 31 intact dogs, ventilated alternately in hyperoxia (fraction of inspired oxygen, FIO2 0.4) and in hypoxia (FIO2 0.1). A sequence of two 5-point Ppa/Q plots at FIO2 0.4 and at FIO2 0.1, consecutively, was performed before and after i.v. administrations of theophylline 10 mg.kg-1 (n = 8) and 25 mg.kg-1 (n = 7), of S 9795 10 mg.kg-1 (n = 8) and of placebo (n = 8). The Ppa/Q plots were rectilinear in all experimental conditions. Over the entire range of Q studied, 2-5 l.min-1.m-2, hypoxia increased Ppa in all animals. Placebo had no effect on these Ppa/Q plots. Theophylline at the lowest dose (plasma levels from 8.4-13.6 micrograms.ml-1) and S 9795 (plasma levels from 3.0-11.2 micrograms.ml-1) did not affect Ppa/Q in hyperoxia or in hypoxia. Theophylline at the highest dose (plasma levels from 17.8-40.4 micrograms.ml-1) reduced hypoxic Ppa at all levels of Q and hyperoxic Ppa at the highest Q, from 3-5 l.min-1.m-2, and inhibited hypoxia-induced increases in Ppa. We conclude that pulmonary vasoconstriction may be preserved after the lowest doses of methylxanthines recommended for the treatment of increased bronchial tone.