While dynamic ocular motor abnormalities (e.g., gaze-evoked nystagmus (GEN), low optokinetic nystagmus (OKN), pursuit and vestibulo-ocular reflex (VOR) gains, and dysmetric saccades) have been shown to be potential biomarkers in spinocerebellar ataxia type 3 (SCA3), the value of static abnormalities (e.g., convergent [esodeviation] and divergent strabismus [exodeviation]) is unknown. Moreover, studies on dynamic abnormalities in SCA3 usually do not take into account the existence of potential abduction-adduction asymmetries in patients with degenerative ataxia. Thirty-eight patients with genetically confirmed SCA3 (24 females; mean age ± SD, 49.8± 12.2 years) and 22 healthy controls (12 females, p = 0.589; mean age ± SD, 50.7± 12.5 years, p = 0.651) underwent clinical and video-oculographic assessment. A p value < 0.002 (between- and within-group analyses) and < 0.001 (correlation analysis) was considered significant. Patients showed larger esodeviation at distance (p < 0.001), became more esodeviated in lateral gaze (p < 0.001), and their near exodeviation correlated with scale for the assessment and rating of ataxia (SARA) score (p = 0.004). Pursuit, OKN, and VOR gains were lower in patients, both for their adducting and abducting components (p < 0.001). Saccades showed higher velocities (p < 0.001), abducting saccades showed lower amplitude (p < 0.001), and adducting saccades tended to show greater vertical bias (p = 0.018) in patients. Abducting saccades showed relatively lower velocity (p < 0.001) and lower amplitude (p = 0.015) than abducting saccades within patients. All dynamic ocular motor abnormalities except saccades correlated with SARA score, CAG repeat number, and/or disease duration (p < 0.001). Static and dynamic ocular motor abnormalities are potential biomarkers in SCA3. SCA3 studies using saccades should take into account the existence of potential abduction-adduction asymmetries.