OBJECTIVE Transient neurological symptoms are frequently observed during the early postoperative period after direct bypass surgery for moyamoya disease. Abnormal signal changes in the cerebral cortex can be seen in postoperative MR images. The purpose of this study was to reveal the radiological features of the "cortical hyperintensity belt (CHB) sign" in postoperative FLAIR images and to verify its relationship to transient neurological events (TNEs) and regional cerebral blood flow (rCBF). METHODS A total of 141 hemispheres in 107 consecutive patients with moyamoya disease who had undergone direct bypass surgery were analyzed. In all cases, FLAIR images were obtained during postoperative days (PODs) 1-3 and during the chronic period (3.2 ± 1.13 months after surgery). The CHB sign was defined as an intraparenchymal high-intensity signal within the cortex of the surgically treated hemisphere with no infarction or hemorrhage present. The territory of the middle cerebral artery was divided into anterior and posterior parts, with the extent of the CHB sign in each part scored as 0 for none; 1 for presence in less than half of the part; and 2 for presence in more than half of the part. The sum of these scores provided the CHB score (0-4). TNEs were defined as reversible neurological deficits detected both objectively and subjectively. The rCBF was measured with SPECT using N-isopropyl-p-[123I]iodoamphetamine before surgery and during PODs 1-3. The rCBF increase ratio was calculated by comparing the pre- and postoperative count activity. RESULTS Cortical hyperintensity belt signs were detected in 112 cases (79.4%) and all disappeared during the chronic period. Although all bypass grafts were anastomosed to the anterior part of the middle cerebral artery territory, CHB signs were much more pronounced in the posterior part (p < 0.0001). TNEs were observed in 86 cases (61.0%). Patients with TNEs showed significantly higher CHB scores than those without (2.31 ± 0.13 vs 1.24 ± 0.16, p < 0.0001). The CHB score, on the other hand, showed no relationship with the rCBF increase ratio (p = 0.775). In addition, the rCBF increase ratio did not differ between those patients with TNEs and those without (1.15 ± 0.033 vs 1.16 ± 0.037, p = 0.978). CONCLUSIONS The findings strongly suggest that the presence of the CHB sign during PODs 1-3 can be a predictor of TNEs after bypass surgery for moyamoya disease. On the other hand, presence of this sign appears to have no direct relationship with the postoperative local hyperperfusion phenomenon. Vasogenic edema can be hypothesized as the pathophysiology of the CHB sign, because the sign was transient and never accompanied by infarction in the present series.