Robotic assistance has improved electrode implantation precision in stereoelectroencephalography (SEEG) for refractory epilepsy patients. We sought to assess the relative safety of the robotic-assisted (RA) procedure compared to the traditional hand-guided one. A systematic search on PubMed, Web of Science, Embase, and Cochrane was performed for studies directly comparing robot-assisted vs. manually guided SEEG to treat refractory epilepsy. The primary outcomes included target point error (TPE), entry point error (EPE), time of implantation of each electrode, operative time, postoperative intracranial hemorrhage, infection, and neurologic deficit. We included 427 patients from 11 studies, of whom 232 (54.3%) underwent robot-assisted surgery and 196 (45.7%) underwent manually guided surgery. The primary endpoint, TPE, was not statistically significant (MD 0.04 mm; 95% CI - 0.21, - 0.29; p = 0.76). Nonetheless, EPE was significantly lower in the intervention group (MD - 0.57 mm; 95% CI - 1.08; - 0.06; p = 0.03). Total operative time was significantly lower in the RA group (MD - 23.66 min; 95% CI - 32.01, - 15.31; p < 0.00001), as well as the individual time of implantation of each electrode (MD - 3.35 min; 95% CI - 3.68, - 3.03; p < 0.00001). Postoperative intracranial hemorrhage did not differ between groups: robotic (9/145; 6.2%) vs. manual (8/139; 5.7%) (RR 0.97; 95% CI 0.40-2.34; p = 0.94). There was no statistically relevant difference in infection (p = 0.4) and postoperative neurological deficit (p = 0.47) incidence between the two groups. In this analysis, there is a potential relevance in the RA procedure when comparing the traditional one, since operative time, time of implantation of each electrode, and EPE were significantly lower in the robotic group. More research is needed to corroborate the superiority of this novel technique.