The purpose of this study is to know clinical factors underlying either a single surgery or repetitive surgeries, required to treat superior oblique muscle palsy. Retrospective review was made on 246 consecutive patients with idiopathic (n = 212) or acquired (n = 34) superior oblique muscle palsy who underwent surgeries in 8 years at one institution. Idiopathic palsy included congenital and decompensated palsies while acquired palsy included traumatic and ischemic palsies. Clinical factors, compared between groups with a single surgery (n = 203) and two or more surgeries (n = 43), were surgical methods, sex, age at surgery, horizontal, vertical, and cyclotorsional deviations, and stereopsis at near fixation. Inferior oblique muscle recession on paretic side was chosen in about 60% of the single-surgery and repetitive-surgery group as an initial surgery, followed by inferior rectus muscle recession on non-paretic side. The age at surgery was significantly older, vertical and cyclotorsional deviations were significantly larger in the repetitive-surgery group, compared with the single-surgery group (P = 0.01, P < 0.001, P = 0.02, Mann–Whitney U-test, respectively). The 95% confidence interval of vertical deviations was 15–17 prism diopters in the single-surgery group and 23–28 prism diopters in the repetitive surgery group. Significant differences in vertical deviations were replicated also in subgroups of patients with either idiopathic or acquired palsy. In conclusions, the 95% confidence interval of vertical deviations, determined by alternate prism and cover test, would be used as a common benchmark for predicting either a single surgery or repetitive surgeries, required to treat idiopathic and acquired superior oblique muscle palsy, in the process of obtaining the informed consent.