Objective To measure obstruction length and height using drug-induced sleep endoscopy (DISE) in obstructive sleep apnea (OSA) patients and to evaluate their association with outcomes of velopharyngeal surgery. Study Design Prospective cohort study. Setting University medical center. Methods Forty-three consecutive patients with OSA were evaluated by DISE using dexmedetomidine. The 2 new parameters, obstruction length (defined as the distance from the most superior point of the collapse to the most inferior point of the collapse) and obstruction height (the distance from the posterior border of the nasal septum to the most proximal point of the collapse), were measured by both DISE and a pressure transducer catheter method before surgery. All of the patients received velopharyngeal surgery, including revised uvulopalatopharyngoplasty with uvula preservation and transpalatal advancement pharyngoplasty. We followed up with all of the patients using polysomnography at least 3 months after surgery. Results Twenty-six (60.5%) patients were responders, and 17 (39.5%) were nonresponders. The mean obstruction length and obstruction height were 1.3 0.5 cm (range, 0.4-2.2 cm) and 3.4 +/- 0.9 cm (range, 1.1-5.0 cm), respectively. Nonresponders had a longer obstruction length and a shorter obstruction value. Multivariate logistic regression analysis revealed that obstruction length >1.4 cm (odds ratio [OR], 0.21; 95% confidence interval [CI], 0.04-0.98; P = .048) and obstruction height 3.2 cm (OR, 9.35; 95% CI, 1.79-48.80; P = .008) were the only independent predictors of velopharyngeal surgery success. Conclusions Accurate measurement of obstruction length and height can be performed with both DISE and a pressure transducer catheter method. The 2 parameters can predict the outcome of velopharyngeal surgery.