The objective of this study was to compare the degree of accuracy in implant positioning and limb alignment offered by two robot-assisted (RA) systems: an image-based robot-assisted (IBRA) versus an imageless robot-assisted (ILRA) system for the treatment of medial knee osteoarthritis with unicompartmental knee arthroplasty (UKA). This retrospective radiographic study included medial UKAs performed between 2011 and 2023. Radiographic measurements taken preoperatively and at 1-year postoperative control visit focusing on hip-knee-ankle angle (HKA), posterior tibial slope (PTS), tibial component coronal alignment relative to Cartier's angle and restoration of proper joint line (JL) height were analyzed. Outliers for postoperative measurements were defined as follows: HKA <175° or >180°, PTS <2° or >8°, >3° or <-3° alterations in Cartier's angle and ±2 mm changes in the height of the joint line. The final sample consisted of 292 medial UKAs: 95 (32.5%) with an IBRA system and 197 (67.5%) with an ILRA system. Implant positioning and limb alignment were more accurate in the group of patients treated with IBRA, HKA (77.9% vs. 67.5%, p = 0.07), PTS (93.7% vs. 82.7%, p = 0.01), restoration of tibial varus relative to Cartier's angle (87.4% vs. 65%, p < 0.001) and restoration of JL height (81.1% vs. 69.5%, p = 0.04). Medial UKA surgery using an IBRA system was associated with a higher degree of accuracy in implant positioning and postoperative limb alignment as compared to an ILRA system. This is a valuable contribution to help communicate the advantages of using this surgical technique and improve its reproducibility. Level III, Retrospective cohort study.