Several choices of instrument systems are available for minimally invasive surgical approaches. There are reports that one alternative, the quadriceps sparing, side-cutting instrumentation, results in diminished implantation accuracy. A total of 108 patients were randomized to undergo TKA either using side-cutting implant instrumentation (Group A) or anterior-posterior mini-incision instrumentation (Group B). All TKAs were operated on with a minimal invasive, mini-midvastus surgical approach. 50% of the TKAs were performed with computer-assistance in each cohort. The radiographic parameters, clinical outcomes and knee scores were evaluated preoperatively and 3 months postoperative. In Group B, we found significantly higher accuracy for the mechanical axis of the limb (range +/-3 degrees: Group A 54% versus Group B 89%, p = <0.001), medial proximal tibial angle (range +/-3 degrees: Group A 85% versus Group B 98%, p = <0.027) and tibial slope (range +/-3 degrees: Group A 59% versus Group B 85%, p = <0.007). The application of the navigation system could only significantly reduce outliers (accuracy >3 degrees) in Group B. Clinical outcomes and knee scores were similar in both groups and were not influenced by computer-assistance. Using the anterior-posterior, mini-incision instruments for minimally invasive TKA will lead to higher implantation accuracy when compared to the quadriceps sparing side-cutting instrumentation. The navigation technique could not compensate for shortcomings of the side-cutting instrumentation. The clinical relevance of this study is that the quadriceps sparing side-cutting instrumentation should not be used for TKA because of unacceptable reduced implantation accuracy.