Whether thrombectomy alone is equally as effective as intravenous alteplase plus thrombectomy remains controversial. We aimed to determine whether thrombectomy alone would be non-inferior to intravenous alteplase plus thrombectomy in patients presenting with acute ischaemic stroke. In this multicentre, randomised, open-label, blinded-outcome trial in Europe and Canada, we recruited patients with stroke due to large vessel occlusion confirmed with CT or magnetic resonance angiography admitted to endovascular centres. Patients were randomly assigned (1:1) via a centralised web server using a deterministic minimisation method to receive stent-retriever thrombectomy alone or intravenous alteplase plus stent-retriever thrombectomy. In both groups, thrombectomy was initiated as fast as possible with any commercially available Solitaire stent-retriever revascularisation device (Medtronic, Irvine, CA, USA). In the combined treatment group, intravenous alteplase (0·9 mg/kg bodyweight, maximum dose 90 mg per patient) was administered as early as possible after randomisation for 60 min with 10% of the calculated dose given as an initial bolus. Personnel assessing the primary outcome were masked to group allocation; patients and treating physicians were not. The primary binary outcome was a score of 2 or less on the modified Rankin scale at 90 days. We assessed the non-inferiority of thrombectomy alone versus intravenous alteplase plus thrombectomy in all randomly assigned and consenting patients using the one-sided lower 95% confidence limit of the Mantel-Haenszel risk difference, with a prespecified non-inferiority margin of 12%. The main safety endpoint was symptomatic intracranial haemorrhage assessed in all randomly assigned and consenting participants. This trial is registered with ClinicalTrials.gov, NCT03192332, and is closed to new participants. Between Nov 29, 2017, and May 7, 2021, 5215 patients were screened and 423 were randomly assigned, of whom 408 (201 thrombectomy alone, 207 intravenous alteplase plus thrombectomy) were included in the primary efficacy analysis. A modified Rankin scale score of 0-2 at 90 days was reached by 114 (57%) of 201 patients assigned to thrombectomy alone and 135 (65%) of 207 patients assigned to intravenous alteplase plus thrombectomy (adjusted risk difference -7·3%, 95% CI -16·6 to 2·1, lower limit of one-sided 95% CI -15·1%, crossing the non-inferiority margin of -12%). Symptomatic intracranial haemorrhage occurred in five (2%) of 201 patients undergoing thrombectomy alone and seven (3%) of 202 patients receiving intravenous alteplase plus thrombectomy (risk difference -1·0%, 95% CI -4·8 to 2·7). Successful reperfusion was less common in patients assigned to thrombectomy alone (182 [91%] of 201 vs 199 [96%] of 207, risk difference -5·1%, 95% CI -10·2 to 0·0, p=0·047). Thrombectomy alone was not shown to be non-inferior to intravenous alteplase plus thrombectomy and resulted in decreased reperfusion rates. These results do not support omitting intravenous alteplase before thrombectomy in eligible patients. Medtronic and University Hospital Bern.< Réduire