Purpose: To determine the incidence of cerebellar atrophy (CA) in patients with intractable temporal lobe epilepsy, whether any clinical factors are significantly associated with CA, whether CA is unilateral or asymmetric and whether this feature has any relationship to the side of epileptogenicity, and whether the presence of CA is related to epilepsy surgery outcome. Methods: We developed a magnetic resonance imaging method of measuring the presurgical volumes of the cerebellar hemispheres of 185 patients who underwent temporal lobectomy for intractable epilepsy and of 80 control subjects. In addition, cerebellar volumes were normalized to the total brain volumes. CA was determined as being present when the measured volume was smaller than two standard deviations from the mean value found in control subjects. Results: Both absolute and normalized cerebellar volumes were found to be significantly reduced in the epilepsy patients compared with the control subjects. Without normalization of the cerebellar volumes, CA was present in 25.9% of the epilepsy patients; with normalization, it was present in only 16.2%. The atrophy was symmetric between the cerebellar hemispheres, and there was no significant difference in volume between the hemisphere ipsilateral and the hemisphere contralateral to the side of the temporal lobectomy. The duration of epilepsy was significantly longer and the age at onset of epilepsy was younger in patients with CA than in those without CA. The presence of CA was not associated with the outcome of temporal lobectomy. Conclusions: CA is symmetric and common in patients with intractable temporal lobe epilepsy. However, the results suggest that the atrophy in one third of patients with CA also proportionately affects the cerebral hemispheres. The duration of epilepsy and the age at onset of epilepsy are associated with the occurrence of CA. Seizure control after temporal lobectomy is not influenced by the presence of CA.