Regression or disappearance of MRI abnormalities is usually observed after treatment of spinal dural arteriovenous fistulae (sDAVF). To assess the correlation between spinal MRI (sMRI) changes with sDAVF exclusion and clinical outcome. Imaging data of patients treated with endovascular embolization for sDAVF between 2007 and 2023 were retrospectively analyzed. Spinal cord edema and perimedullary flow voids at baseline and 3-months' follow-up were compared between patients with and without sDAVF persistent occlusion and clinical improvement on the Aminoff and Logue Scale. Twenty-five patients were included in this study. At 3-months' follow-up, regression of spinal cord edema was significantly associated with sDAVF persistent occlusion (P=0.038). The combination of edema and flow voids regression was significantly associated with higher odds of a cured sDAVF (P<0.001) and clinical improvement (P<0.01). Improvement in the combination of the above-mentioned sMRI signs presented high sensitivity (100% (95% CI 78.20%-100%)) and negative predictive value (100% (95% CI 47.82%-100%)) for the detection of sDAVF cure compared with the criterion standard (digital subtraction angiography (DSA)). Patients with both spinal cord edema and flow voids regression at 3 months were more likely to present with a persistent occlusion of sDAVF and clinical improvement after endovascular embolization. Patients without sMRI improvement should be referred for DSA to seek recurrence of sDAVF.