We hypothesized that bolus injections of lipid soluble chemotherapeutic drugs during transient cerebral hypoperfusion could significantly boost regional drug delivery. In the first two groups of New Zealand White rabbits we measured brain tissue carmustine concentrations after intravenous infusion, intraarterial infusion with normal perfusion, and after intraarterial injections during transient cerebral hypoperfusion. In the third group of animals we assessed the safety of the technique by assessing electroencephalographic changes for 6 h after flow arrest carmustine administration and subsequent histological examination. The brain tissue carmustine concentrations were fivefold to sevenfold higher when the drug was injected during cerebral hypoperfusion compared to a conventional intracarotid infusion (68.4 +/- 24.5 vs. 14.2 +/- 8.3 microg/g, n = 5 each, respectively, P < 0.0001). The brain tissue carmustine concentrations (y) were a linear function of the bolus dose (x) injected during cerebral hypoperfusion, y = 10.4 x x - 21 (R = 0.84, P < 0.001). Stable EEGs were recorded several hours after flow arrest carmustine exposure and histological examinations did not reveal any gross evidence of cerebral injury. Transient cerebral hypoperfusion during intraarterial bolus injection of carmustine significantly increases drug delivery. Clinical techniques that decrease CBF, such as, transient arterial occlusion by balloon tipped catheters, hyperventilation, hypothermia, induced hypotension, or transient circulatory arrest, could enhance intraarterial drug delivery to the brain. We believe that the mechanisms for improved drug delivery is the decrease in drug dilution by reduced or absent blood flow, decreased protein binding and a longer time for high concentrations of free drugs to transit through the blood brain barrier.