OBJECTIVE: To design and evaluate a new vestibular implant and surgical procedure that should reach correct electrode placement in 95% of patients in silico. DESIGN: Computational anatomy driven implant and surgery design study. SETTING: Tertiary referral center. PARTICIPANTS: The population comprised 81 patients that had undergone a CT scan of the Mastoid region in the Maastricht University Medical Center. The population was subdivided in a vestibular implant eligible group (28) and a control group (53) without known vestibular loss. INTERVENTIONS: Canal lengths and relationships between landmarks were calculated for every patient. The relationships in group-anatomy were used to model a fenestration site on all three semicircular canals. Each patient's simulated individual distance from the fenestration site to the ampulla was calculated and compared with the populations average to determine if placement would be successful. MAIN OUTCOME MEASURES: Lengths of the semicircular canals, distances from fenestration site to ampulla (intralabyrinthine electrode length), and rate of successful electrode placement (robustness). RESULTS: The canal lengths for the lateral, posterior, and superior canal were respectively 12.1 mm ± 1.07, 18.8 mm ± 1.62, and 17.5 mm ± 1.23, the distances from electrode fenestration site to the ampulla were respectively 3.73 mm ± 0.53, 9.02 mm ± 0.90, and 5.31 mm ± 0.73 and electrode insertions were successful for each respective semicircular canal in 92.6%, 66.7%, and 86.4% of insertions in silico. The implant electrode was subsequently revised to include two more electrodes per lead, resulting in a robustness of 100%. CONCLUSIONS: The computational anatomy approach can be used to design and test surgical procedures. With small changes in electrode design, the proposed surgical procedure's target robustness was reached.