Randomized controlled trials (RCTs) have demonstrated powerful efficacy of endovascular thrombectomy (EVT) for large vessel occlusion in the anterior circulation. The effect of EVT for acute basilar artery occlusion (BAO) in the posterior circulation remains unproven. Here, we highlight the latest findings of observational studies and RCTs of EVT for BAO, with a focus on the predictors of functional outcomes, the limitations of recent RCTs, and critical thinking on future study design. Pooled data from large retrospective studies showed 36.4% favorable outcome at 3 months and 4.6% symptomatic intracranial hemorrhage (sICH). Multivariate logistic regression analysis revealed that higher baseline NIHSS score, pc-ASPECTS < 8, extensive baseline infarction, large pontine infarct, and sICH were independent predictors of poor outcome. Two recent randomized trial BEST (Endovascular treatment vs. standard medical treatment for vertebrobasilar artery occlusion) and BASICS (Basilar Artery International Cooperation Study) failed to demonstrate significant benefit of EVT within 6 or 8 h after stroke symptom onset. The limitations of these studies include slow enrollment, selection bias, high crossover rate, and inclusion of patients with mild deficit. To improve enrollment and minimize risk of diluting the overall treatment effect, futile recanalization and re-occlusion, optimal inclusion/exclusion criteria, including enrollment within 24 h of last known well, NIHSS score ≥ 10, pc-ASPECTS ≥ 8, no large pontine infarct, and the use of rescue therapy for underlying atherosclerotic stenosis, should be considered for future clinical trials.