Background: Surgical robots have been appearing in operating rooms over the past decade, and neurosurgery has been one of the pioneers in this area. In neurosurgery, the clinical use of robots has been limited to stereotactic procedures and endoscopic manoeuvres, although the brain is a unique organ and well-suited for robotic application. The aim of this study was to assess the ability of our vision-guided robotic system to perform basic neurosurgical procedures. Methods: The study was divided into two parts: bone drilling and endoscopic manoeuvres. The robotic system was instructed to recognise targets on artificial skull models placed in different positions (supine, lateral, sitting, and prone) and to make burr holes. A total of 10 selected burr holes were used to assess the capability of the robot to insert an endoscope. Results: The accuracy ranged 0.1–1.0 mm with repeatability ranged 0.03–0.92 mm. Conclusion: Generally, the present robotic system is able to perform the surgical tasks. However, further study is needed to refine the robotic system, including the safety mechanisms.