BACKGROUND Thiopental sometimes causes bronchospasm during induction of anaesthesia. In addition, we have reported previously that thiopental produced transient bronchospasm, which was blocked by atropine pretreatment, and worsened histamine-induced bronchoconstriction in dogs. Previous in vitro reports suggest that synthesis of contractile cyclooxygenase products, such as thromboxane A(2), may be involved in the mechanism of bronchospasm. However, the in vivo spastic effects have not been defined comprehensively. METHODS Twenty-seven mongrel dogs were anaesthetized with pentobarbital. Bronchoconstriction was elicited with methacholine (0.5 microg kg(-1)+5.0 microg kg(-1) min(-1); Mch group, n=7) or serotonin (10 microg kg(-1)+1 mg kg(-1) h(-1); 5HT group, n=20), and assessed as percentage changes in bronchial cross-sectional area (BCA, basal=100%) using a bronchoscope. In the 5HT group, dogs were subdivided into four groups of five each: S-5HT, I-5HT, 5HT-S and 5HT-A. In the S-5HT and I-5HT groups, 30 min before serotonin infusion dogs were given saline and indomethacin respectively at 5 mg kg(-1) i.v. In all groups, 30 min after bronchoconstrictor infusion started, dogs were given thiopental at doses between 0 (saline) and 20 mg kg(-1). In the 5HT-S and 5HT-A groups, dogs were given saline or atropine 0.2 mg kg(-1) i.v. 5 min after thiopental 20 mg kg(-1). RESULTS Methacholine and serotonin reduced BCA by about 50 and 40% respectively. Thiopental 20 mg kg(-1) increased and decreased BCA by about 20 and 10% in the Mch and 5HT groups respectively. Indomethacin and atropine did not attenuate the potentiation of serotonin bronchoconstriction produced by thiopental. CONCLUSIONS The present study indicates that thiopental may attenuate or worsen bronchoconstriction induced by muscarinic or serotonin receptor stimulation, respectively. The synthesis of contractile cyclooxygenase products and cholinergic stimulation may not be involved in the contractile effect of thiopental on serotonin bronchoconstriction.