Objective To investigate the prognostic factors for poor postoperative outcomes of patients with dural arteriovenous fistulas (DAVFs). Methods The clinical data of 153 patients with DAVFs, who were treated in our institute over the past 6 years, were retrospectively analyzed. The clinical manifestations, angioarchitecture, treatment methods, neuroradiological results, and clinical outcomes were collected for Kaplan-Meier estimation and Cox regression analysis. Results The patients were followed up for a mean of (38.1±16.3) months. Twenty-one patients showed modified Rankin scale (mRS) 3-6 during recent follow-up; 4 patients died of postoperative complications, another 8 patients died during follow-up, and 9 patients presented with permanent neurologic defects. Kaplan-Meier estimation suggested that patients with poor preoperative neurologic status (mRS≥3), venous sinus thrombosis, combined arteriovenous approach embolization, partial obliteration, parital obliteration with cortical venous reflux (CVR), and postoperative restriction of dominant venous sinuses tended to have poor clinical outcomes. Cox regression analysis demonstrated that poor preoperative neurologic status (mRS≥3) (P=0.018), partial obliteration with CVR (P=0.001), and postoperative restriction of dominant venous sinuses (P=0.000 1) were risk factors predicting poor postoperative outcomes. Conclusion Poor preoperative neurologic status, partial obliteration with CVR, and postoperative restriction of dominant venous sinuses are the independent risk factors for poor clinical outcomes of DAVFs.