Treatments developed in small-animal models of stroke are often not translatable to the clinic. In this protocol, a blood clot in a microcatheter is introduced into the middle cerebral artery of rhesus monkeys; controlled thrombolysis is possible. Over decades of research into the treatment of stroke, nearly all attempts to translate experimental treatments from discovery in cells and rodents to use in humans have failed. The prevailing belief is that it might be necessary to pretest pharmacological neuroprotection in higher-order brains, especially those of nonhuman primates (NHPs). Over the past few years, chemical thrombolysis and mechanical thrombectomy have been established as the standard of care for ischemic stroke in patients. The spotlight is now shifting towards emphasizing both focal ischemia and subsequent reperfusion in developing a clinically relevant stroke model in NHPs. This protocol describes an embolic model of middle cerebral artery occlusion in adult rhesus monkeys. An autologous clot is combined with a microcatheter or microwire through endovascular procedures, and reperfusion is achieved through local intra-artery thrombolysis with tissue plasminogen activator. These NHP models formed relatively stable infarct sizes, delivered predictable reperfusion and survival outcomes, and recapitulated key characteristics of patients with ischemic stroke as observed on MRI images and behavioral assays. Importantly, treated animals could survive 30 d after the surgery for post-stroke neurologic deficit analyses. Thus far, this model has been used in several translational studies. Here we describe in detail the teamwork necessary for developing stroke models of NHPs, including the preoperation preparations, endovascular surgery, postoperation management and histopathological analysis. The model can be established by the following procedures over a 45-d period, including preparation steps (14 d), endovascular operation (1 d) and evaluation steps (30 d).