Robot-assisted percutaneous insertion of pedicle screws is a recent technique demonstrating high accuracy. The optimal treatment for spondylodiscitis is still a matter of debate. We performed a retrospective cohort study on surgical patients treated with pedicle screw/rod placement alone without the application of intervertebral cages. In this collective, we compare conventional open to a further minimalized percutaneous robot-assisted spinal instrumentation, avoiding a direct contact of implants and infectious focus. 90 records and CT scans of patients treated by dorsal transpedicular instrumentation of the infected segments with and without decompression and antibiotic therapy were analysed for clinical and radiological outcome parameters. 24 patients were treated by free-hand fluoroscopy-guided surgery (121 screws), and 66 patients were treated by percutaneous robot-assisted spinal instrumentation (341 screws). Accurate screw placement was confirmed in 90 % of robot-assisted and 73.5 % of free-hand placed screws. Implant revision due to misplacement was necessary in 4.95 % of the free-hand group compared to 0.58 % in the robot-assisted group. The average intraoperative X-ray exposure per case was 0.94 ± 1.04 min in the free-hand group vs. 0.4 ± 0.16 min in the percutaneous group (p = 0.000). Intraoperative adverse events were observed in 12.5 % of free-hand placed pedicle screws and 6.1 % of robot robot-assisted screws. The mean postoperative hospital stay in the free-hand group was 18.1 ± 12.9 days, and in percutaneous group, 13.8 ± 5.6 days (p = 0.012). This study demonstrates that the robot-guided insertion of pedicle screws is a safe and effective procedure in lumbar and thoracic spondylodiscitis with higher accuracy of implant placement, lower radiation dose, and decreased complication rates. Percutaneous spinal dorsal instrumentation seems to be sufficient to treat lumbar and thoracic spondylodiscitis.