Classical conditioning of the eyeblink response is a form of motor learning that is controlled by the intermediate cerebellum and related brainstem structures. A major component of this neural network is the inferior olive (IO), which is the exclusive source of cerebellar climbing fibers. The functional role of the IO in eyeblink conditioning circuits is not completely understood. Pertinent to eyeblink conditioning, there are two competing concepts explaining IO function. The cerebellar learning hypothesis assumes that, by conveying the unconditioned stimulus (US) input, the IO is supplying the cerebellum with a "teaching" signal required for the acquisition of conditioned responses (CR). This hypothesis predicts that blocking US input should lead to the extinction of CRs. An alternate concept assumes that the IO regulates the tonic activity of the neural network, and predicts that inactivation of the IO would produce dysfunction of the cerebellum. To examine these hypotheses, we designed a series of studies probing the behavioral and electrophysiological effects of blocking US sensory input to the IO. The unique feature of these experiments is that they are the first to document effects of IO manipulations on the neuronal activity in the cerebellum.;In the first series of experiments, we found that blocking glutamatergic sensory inputs to the IO or inactivating the IO gradually abolished CRs. In additional tests we determined that the gradual disappearance of CRs was related to drug diffusion rather than to CR extinction. Since these behavioral effects were paralleled by a dramatic suppression of interposed nucleus (IN) neuronal activity, we concluded that inactivating the rostral parts of the IO complex abolishes CRs by producing a tonic dysfunction of cerebellar eyeblink conditioning circuits.;In the second series of experiments, we tested a recent suggestion that selectively blocking AMPA/kainate receptors in the IO blocks the teaching signal to the cerebellum without affecting the tonic state of the cerebellum. In this pivotal experiment, we found that blocking AMPA/kainate receptors in the IO abolishes CRs by suppressing both the tonic and CR-related neuronal activity in the IN and that the delayed (extinction-like) behavioral effects of NBQX most likely stem from the gradual diffusion of injected drug.;The first two experiments significantly weakened the cerebellar learning hypothesis because they demonstrated that the physiological mechanisms explaining the behavioral effects of common IO manipulations is inconsistent with this notion. We proposed that the ultimate test of the cerebellar learning hypothesis would be blocking IO sensory inputs without upsetting the tonic activity in the cerebellum. To achieve this goal, we initiated a third group of studies to develop a new pharmacological approach that would counter-balance the tonic effects of glutamate blockers. We examined two drug candidates, harmaline and gabazine, for their potential to elevate IO spontaneous activity. In these preliminary studies, we found that the GABA-A antagonist gabazine could recover CRs and IN neuronal activity that had been suppressed by the prior blocking of glutamate receptors with DGG. These ongoing experiments suggest that a near normal level of IO firing is critical for CR performance, and that US signals via the IO are most likely not required for the maintenance of learned responses.