Background: Medium vessel occlusions (MeVOs) are an increasingly recognized but heterogeneous target for endovascular therapy (EVT). This study aims to compare primary MeVO, rescue MeVO, and large vessel occlusion (LVO) thrombectomy cases to identify which MeVO subtypes derive a meaningful benefit from EVT under appropriate safety conditions. Methods: We retrospectively analyzed a multicenter registry of patients undergoing EVT for acute ischemic stroke. MeVO was defined as the occlusion of the A1-A3, M2-M3, P1-P3, fetal PCA, or PICA segments and classified as primary or rescue. Clinical outcomes were assessed by NIHSS score at baseline, discharge, and 90 days; functional outcome by the modified Rankin scale (mRS); and reperfusion by modified thrombolysis in cerebral infarction (mTICI). Safety endpoints included intracranial hemorrhage and mortality. Results: Among 603 EVT patients, 202 (33.5%) had MeVO. Compared to LVO, MeVO patients were older and had more prior strokes but achieved similar reperfusion and safety outcomes. At 90 days, mRS distribution differed, with MeVO showing more mRS 2 and LVO more mRS 1, while higher-disability strata were comparable. Within MeVO, 119 (58.9%) were primary and 83 (41.1%) rescue occlusions. Rescue MeVO patients presented with higher baseline severity (NIHSS score of 19 vs. 18) and, despite similar reperfusion, experienced worse 90-day outcomes and higher mortality (21.7% vs. 0.8%). Conclusions: EVT for primary MeVO is feasible, effective, and safe, whereas rescue MeVO is associated with poor functional outcomes and markedly higher mortality. These findings highlight rescue MeVO as a distinct phenotype and support a selective approach prioritizing disabling syndromes, proximal/dominant branch occlusions, and IVT non-response.