Background: Ankle arthroscopy has evolved as a diagnostic and therapeutic tool. Traditionally, it is performed with traction because of the tight ankle joint space. Original traction techniques were invasive but have progressed to the commonly used noninvasive modalities. Recent reports have suggested traction may be unnecessary. The purpose of this study was to compare prospectively ankle arthroscopy with and without traction in terms of ease of visualization of anatomic structures according to the Ferkel's ankle arthroscopy criteria. Methods: Under Institutional Review Board approval, 103 patients received ankle arthroscopies, first with noninvasive traction and subsequently without traction. An independent observer scored each arthroscopy based on the 21-point Ferkel's criteria. For each structure adequately visualized, one point was given. Inadequate visualization was defined as excessive force or as visible cartilage scuffing. The maximum possible score was 21 points and the lowest was zero. Fifty-five females and 48 males, average age 35 (range, 16 to 71) years, participated in the study. Results: Noninvasive traction facilitated visualization of all structures of the anterior ankle in more than 90% of cases except for the anterior compartment and lateral gutters, which were better visualized without traction with the ankle in dorsiflexion. No difference was seen when visualizing the talus. Traction arthroscopy performed better when evaluating the central and posterior ankle. The mean score difference was statistically significant, 11.2 versus 18.5 points, favoring noninvasive traction arthroscopy. The complication rate was 4%. Conclusion: Noninvasive traction facilitated complete ankle arthroscopy. Dorsiflexion improved visualization of the anterior compartment and lateral ankle gutter. The authors recommend noninvasive traction when performing ankle arthroscopy. Level of Evidence: II, Diagnostic, Prospective Study