BACKGROUND: Anatomic hemispherectomy is an effective surgical treatment for patients with hemispherical intractable epilepsy. Different degrees of brain shifting have been observed, but whether these shifts can predict motor function recovery is unknown. The aim of this study was to analyze the correlation between brain shift ratios of different brain areas and motor function before and after surgery. METHODS: Medical records and magnetic resonance imaging of 23 patients who underwent anatomic hemispherectomy from 2006 to 2013 at a single center were retrospectively reviewed. Proximal and distal muscle strengths of both arms and legs were measured before and after surgery to determine motor function outcomes. Brain shift ratios of frontal lobe, temporal lobe, parieto- occipital lobe, thalamus, brainstem, and cerebellum were measured before and after surgery to test which shifts could effectively predict motor function outcomes. RESULTS: Fifteen patients (65.2%) showed different degrees of presurgical brain shifting. Brain shift ratios of all measured brain areas were generally increased after anatomic hemispherectomy. After surgery, 13 patients (56.5%) exhibited improved proximal muscle strength, whereas 10 (43.5%) in distal. Significant correlations were found only between muscle strength improvements of distal arms or legs and presurgical brain shift ratios of thalamus or brainstem (all P < 0.05). Distal muscle strength improvements also correlated with age at seizure onset. CONCLUSIONS: Patients with hemispherical intractable epilepsy with larger presurgical shifts of thalamus and brainstem exhibited improved muscle strength, especially in distal muscles, after anatomic hemispherectomy. This result was more likely in patients who were older at the time of seizure onset. These presurgical shifts of thalamus and brainstem may be used for predicting motor function recovery after hemispherectomy for a subset of patients, which is beneficial for surgical planning.