Focal gastric mucosal blood flow as measured by the hydrogen gas clearance method was compared with total gastric blood flow as determined by venous outflow in an isolated segment of canine stomach before, during, and after exposure to aspirin. Despite an increase in total gastric blood flow from 10.8 +/- 1.6 ml/min per chamber to 17.4 +/- 1.9 ml/min per chamber and mucosal blood flow at nonulcerated sites from 29.5 +/- 4.3 ml/min per 100 g to 83 +/- 14.4 ml/min per 100 g, mucosal blood flow at the site of aspirin-induced ulceration was significantly reduced from 29.5 +/- 4.3 ml/min per 100 g to 12.5 +/- 2.5 ml/min per 100 g. After the removal of aspirin, mucosal blood flow returned to control levels. Such a redistribution of mucosal blood flow in response to aspirin is consistent with the localized nature of acute aspirin-induced injury. The findings also explain the inability of previous methods measuring blood flow of the entire stomach to demonstrate mucosal ischemia during such injury.