Operative manipulation of blood vessels might lead to spasm, thereby destroying the endothelial cell function: the spasm can be prevented by the vasodilator papaverine. To study if this was mediated via the prostanoid pathway the following investigation was undertaken: canine jugular veins and carotid arteries were dissected with or without papaverine. Vessel segments were then perfused with Hank's balanced salt solution for five times 15 min. Prostacyclin was measured as the stable degradation product 6-keto-PGF1 alpha and thromboxane as TXB2, by radioimmunoassay. Control arterial segments' 6-keto-PGF1 alpha release was initially 129.5 + 20.1 pg/mm2/15 min, and 29.7 + 10.4 after 60 min (p less than 0.05 vs initial value) and responded to arachidonic acid (AA) with an increase to 139.2 +/- 23.1 pg/mm2/15 min (p less than 0.05). Segments treated with papaverine had the same release as the controls. In venous segments there was a lower initial release (p less than 0.05) from segments given papaverine than from controls, but this was more likely an effect of papaverine on the assay. There was no difference in release of prostacyclin from segments given papaverine in the perfusate compared to controls when using 125I tracer. When using 3H tracer including absorption of free antigen to dextran coated charcoal, papaverine displaced the free tracer giving artificially low values. There was no effect of papaverine given intraoperatively on the TXB2 release, neither from arteries nor from veins. In another experiment the vessel wall tension was examined and the cyclooxygenase inhibitor diclofenac did not inhibit the vasodilating effect of papaverine.(ABSTRACT TRUNCATED AT 250 WORDS)