We used transgenic mice with Purkinje cell dysfunction (PO3 line) to study the role of these neurons in the increase in cerebellar blood flow (BF_crb) produced by stimulation of the cerebellar parallel fibers (PF). Mice (age 8-10 wk) were anesthetized (halothane) and artificially ventilated. Arterial pressure and end-tidal CO₂ were monitored continuously. Arterial blood gases were measured. The PF were stimulated electrically (100 µA, 30 Hz; 40 s), and the increases in BF_crb were monitored by a laser-Doppler flow probe. First, we characterized the increases in BF_crb and the field potentials produced by PF stimulation in normal mice. PF stimulation evoked the typical field potentials and increased BF_crb by 60 ± 4% (100 µA, 30 Hz; n = 10). The increases in BF_crb were attenuated by the broad-spectrum glutamate receptor antagonist kynurenate (84 ± 3%; P < 0.05 analysis of variance; n = 5), by the DL--amino-3-hydroxy-5-methylisoxazole-4-propionic acid receptor antagonist 2,3-dihydroxy-6-nitro-7-sulfamoyl-benzo(f)quinoxaline (62 ± 6%; P < 0.05; n = 5), and by the nitric oxide synthase inhibitor N-nitro-L-arginine (46 ± 7%; P < 0.05; n = 5). In PO3 transgenic mice, the increases in BF_crb produced by PF stimulation were reduced (P < 0.001) at every stimulus intensity and frequency tested (residual increase at 100 µA, 30 Hz: 19 ± 2%; n = 6). The field potentials evoked by PF stimulation also were abnormal in that they lacked the late negative wave (n = 6), a finding consistent with lack of depolarization of Purkinje cells. The residual flow response in the transgenics was abolished by N-nitro-L-arginine (n = 5; P > 0.05). Ultrastructural studies showed that the density of PF-Purkinje cell synapses is reduced in PO3 mice, whereas the morphology of molecular layer interneurons (stellate cells) is normal. The findings suggest that Purkinje cells are responsible for a sizable component of the flow response whereas molecular layer interneurons mediate the remainder of the response. The study provides evidence that mouse mutants with spontaneous or genetically engineered cerebellar abnormalities could be useful to study the cellular and molecular correlates of functional hyperemia in the central nervous system.