Introduction The goal of this investigation was to compare fusion of sequential cone beam computerized tomography (CT) volumes to the gold standard (fiducial registration) in order to be able to analyze clinical cochlear implant (CI) migration with high accuracy in three dimensions. Materials and methods Paired cone beam CT volumes were performed on five human cadaver temporal bones and one human subject. These volumes were fused using 3D Slicer 4 and BRAINSFit software. Using a gold standard fiducial technique, the accuracy, robustness, and performance time of the fusion process were assessed. Results This proposed fusion protocol achieves a subvoxel median Euclidean distance of 0.05 mm in human cadaver temporal bones and 0.16 mm (mean) when applied to the described in vivo human synthetic data set in over 95% of all fusions. Performance times are <2 min. Conclusion Here, a new and validated method based on existing techniques is described, which could be used to accurately quantify migration of CI electrodes.