Background: A large infarct and expanding cerebral edema (CED) due to a middle cerebral artery occlusion confers a 70% mortality unless treated surgically. There is still conflicting evidence whether reperfusion is associated with a lower risk for CED in acute ischemic stroke. Aim: To investigate the association of reperfusion with development of early CED after stroke thrombectomy. Methods: From the SITS-International Stroke Thrombectomy Registry, we selected patients with occlusion of the intracranial internal carotid or middle cerebral artery (M1 or M2). Successful reperfusion was defined as mTICI ⩾ 2b. Primary outcome was moderate or severe CED, defined as focal brain swelling ⩾1/3 of the hemisphere on imaging scans at 24 h. We used regression methods while adjusting for baseline variables. Effect modification by severe early neurological deficits, as indicators of large infarct at baseline and at 24 h, were explored. Results: In total, 4640 patients, median age 70 years and median National Institutes of Health Stroke Score (NIHSS) 16, were included. Of these, 86% had successful reperfusion. Moderate or severe CED was less frequent among patients who had reperfusion compared to patients without reperfusion: 12.5% versus 29.6%, p < 0.05, crude risk ratio (RR) 0.42 (95% confidence interval (CI): 0.37–0.49), and adjusted RR 0.50 (95% CI: 0.44-0.57). Analysis of effect modification indicated that severe neurological deficits weakened the association between reperfusion and lower risk of CED. The RR reduction was less favorable in patients with severe neurological deficits, defined as NIHSS score 15 or more at baseline and at 24 h, used as an indicator for larger infarction. Conclusion: In patients with large artery anterior circulation occlusion stroke who underwent thrombectomy, successful reperfusion was associated with approximately 50% lower risk for early CED. Severe neurological deficit at baseline seems to be a predictor for moderate or severe CED also in patients with successful reperfusion by thrombectomy.