Pial arterioles of mice were suffused in situ with terf-butyl hydroperoxide (TBHP). This agent has been reported to stimulate synthesis or release of the constrictor, thromboxane. In the present study we observed pial arterioles by in vivo microscopy. Locally applied TBHP produced dose-dependent constriction, significantly inhibited by each of three drugs: acet visa] icy lie acid, OKY-046, and SQ-29548. These drugs are respectively a cyclooxygenase inhibitor, a thromboxane synthetase inhibitor, and a thromboxane receptor blocker. Since each acts by a different mechanism to interfere with thromboxane-mediated responses and each inhibited the contractile response to TBHP, thromboxane appears to be a mediator of this response. Platelet aggregation was not seen after local application of TBHP, and the dwell time of platelets at the site of TBHP application is less than 1 second. Thus, platelets are an unlikely source of the thromboxane mediating the local constriction. It is much more likely that the source of thromboxane is either the wall of the pial vessels or the underlying brain and/or its vessels. These data do not distinguish between the latter two possibilities, but if this suggestion is correct, then the data show for the first time that thromboxane can be released from normal brain and/or brain vessels in amounts sufficient to cause arteriolar constriction. (Stroke 1987; 18:195-199) T HROMBOXANE has long been known to be a potent constrictor of cerebral arteries and other vessels. '~3 However, there has been no published evidence to suggest that cerebral blood vessels can be constricted by thromboxane released from the vessels themselves and/or adjacent brain. Rather, demonstrations of constriction have involved in vitro application of thromboxane or its analogs.14 ' 5 A similar situation is encountered in the literature for other vascular beds, probably because of the difficulty of selectively releasing endogenous thromboxane from the vessels or organ in question. However, Gurtner et al 6 found that tert-bnty\ hydroperoxide (TBHP) caused marked constriction of the perfused pulmonary vasculature, which was associated with the appearance of increased levels of thromboxane in the effluent. The elevated thromboxane level and the associated vasoconstriction were prevented by blocking cyclooxygenase, which was interpreted as indicating that TBHP stimulates cyclooxygenase-dependent arachidonate metabolism with a resultant increase in the production of eicosanoids like thromboxane. However, thromboxane production was increased at least eightfold while prostacyclin production was increased only threefold by the TBHP. Thus, TBHP might in addi- Received June 15, 1986; accepted August 28, 1986. tion selectively stimulate thromboxane synthetase by acting on cyclooxygenase-dependent precursors of thromboxane. In either case the thromboxane produced by TBHP caused significant constriction. The source of the thromboxane was thought to be lung "parenchyma" 6 ; however, the possibility of thromboxane released from v...