In 53 patients with chronic occlusive cerebrovascular diseases, we evaluated cerebral perfusion reserve. Cerebral perfusion reserve was evaluated by the change in cerebral blood flow (CBF) at rest and the CBF after acetazolamide (Diamox) as measured by the 133Xe gas inhalation SPECT method. Crossed cerebellar diaschisis (CCD) was evaluated semiquantitatively by the count ratio of the cerebellar cortex using 99mTc-HMPAO SPECT. Thirty-two (49%) of the 65 affected lesions in 53 patients had shown a decreased response to Diamox. In the cases of decreased response to Diamox, the response improved significantly after surgery. However, CBF did not improve as much as the response. In the 40 nonaffected sides, 21 (53%) showed decreased CBF at rest but good response to Diamox. At rest, no difference of the CBF ratio (affected/contralateral nonaffected cerebral cortex) was observed between the patients with CCD (6 pts) and those without CCD (8 pts). After Diamox, however, the CBF ratio of the patients with CCD was significantly higher than that without CCD. The change of the CBF ratio before and after Diamox correlated significantly with the degree of CCD (r = -0.794, p < 0.01). Diamox was useful for evaluating the cerebral perfusion reserve to indicate surgery in patients with chronic cerebrovascular diseases. Perhaps CCD could be a useful index in the differentiation of the decreased CBF caused by reduced perfusion pressure from that caused by reduced metabolic demand because CCD had a close relationship with the cerebral perfusion reserve.