PURPOSE: Stereoelectroencephalography (SEEG) is the standard for invasive investigations in epilepsy surgery. Our aim was to investigate if similar precision and low complication rate can be achieved with optimized frameless navigation as with frame-based or dedicated stereotactic SEEG robot. METHODS: We compared five different implantation techniques assessing entry, target errors and complications in 53 SEEGs from 50 patients: Group 1 - surface registration and Vertek probe, Group 2 - rigid registration with conventional CT and Vertek probe, Group 3 - rigid registration and Vertek probe, Group 4 - rigid registration and Autoguide, Group 5 - rigid, sterile registration and Autoguide. Analysis was done using random effects linear modelling to calculate improvement in percent using Group 1 as a reference, p < 0.001 was considered significant. RESULTS: Mean patient age at implantation was 23 years (range 4-46 years) and mean number of implanted electrodes per patient were 11 (range 3-15). Accuracy data was available for 36 SEEG implantations (419 electrodes). The median entry/target errors were (mm): Group 1:4.6/4.3; Group 2:1.8/2.3; Group 3:0.9/1.5; Group 4:1.1/1.2; Group 5:0/0.7. Improvement of accuracy for entry error was 38% for Group 2 (p = 0.004), 47% for Group 3 (p < 0.001), 50% for Group 4 (p < 0.001), and 72% for Group 5 (p < 0.001). Improvement of accuracy for target error was 17% for Group 2 (p = 0.17), 22% for Group 3 (p < 0.001), 35% for Group 4 (p < 0.001), and 51% for Group 5 (p < 0.001). Complications (hemorrhage, edema, headache) occurred in 7/53 SEEGs, none of these led to permanent deficit. 40/53 investigations resulted in an epilepsy surgery procedure. CONCLUSION: High precision and low complication rate in SEEG implantation can be achieved with frameless navigation using rigid, sterile registration.