Previous studies have shown a complete plasma perfusion of all capillaries in the rat brain under normal physiological conditions. This raises the question under which experimental conditions nonperfused capillaries may show up in the brain. Two experimental models were investigated in rats. 1) Reduced cerebral blood flow (CBF) during incomplete forebrain ischemia: hemorrhagic hypotension was maintained for 30 min at a mean arterial blood pressure of 41 mmHg. During the final 5 min of hypotension both carotid arteries were ligated. 2) Reperfusion after incomplete forebrain ischemia: reperfusion lasted for 4 h after either 15 or 30 min of incomplete forebrain ischemia. Under both experimental conditions, the density of the existing as well as the plasma-perfused brain capillary network was quantified using fluorescent double staining. Local CBF was measured during incomplete forebrain ischemia using the quantitative autoradiographic 4-iodo-[N-methyl-14C]antipyrine technique. The results showed a decrease in CBF during incomplete forebrain ischemia, which amounted up to 94%. Whereas normotensive control animals showed a complete staining of all capillaries within 5 s after the intravenous injection of Evans blue, this period of time was increased to 10 s during incomplete forebrain ischemia, indicating a delayed capillary perfusion. Four hours of reperfusion after 15 min of incomplete forebrain ischemia resulted in a complete capillary staining, whereas reperfusion after 30 min of ischemia was followed by intracerebral bleedings and a few nonperfused capillary areas (circulation time of Evans blue: 10 s).(ABSTRACT TRUNCATED AT 250 WORDS)