Although intravenous thrombolysis increases the probability of a goodfunctional outcome in carefully selected patients with acute ischemicstroke, a substantial proportion of patients who receive thrombolysis do nothave a good outcome. Several recent trials of mechanical thrombectomy appearto indicate that this treatment may be superior to thrombolysis. Wetherefore conducted a systematic review and meta-analysis to evaluate theclinical effectiveness and safety of new-generation mechanical thrombectomydevices with intravenous thrombolysis (if eligible) compared withintravenous thrombolysis (if eligible) in patients with acute ischemicstroke caused by a proximal intracranial occlusion. We systematicallysearched seven databases for randomized controlled trials published betweenJanuary 2005 and March 2015 comparing stent retrievers or thromboaspirationdevices with best medical therapy (with or without intravenous thrombolysis)in adults with acute ischemic stroke. We assessed risk of bias and overallquality of the included trials. We combined the data using a fixed or randomeffects meta-analysis, where appropriate. We identified 1579 studies; ofthese, we evaluated 122 full-text papers and included five randomizedcontrol trials (n=1287). Compared with patients treated medically, patientswho received mechanical thrombectomy were more likely to be functionallyindependent as measured by a modified Rankin score of 0-2 (odds ratio, 2.39;95% confidence interval, 1.88-3.04; I2=0%). This finding wasrobust to subgroup analysis. Mortality and symptomatic intracerebralhemorrhage were not significantly different between the two groups.Mechanical thrombectomy significantly improves functional independence inappropriately selected patients with acute ischemic stroke.