21 randomized controlled trials analyzing a total of 1713 knees were included in this meta-analysis comparing radiographic alignment and functional outcome between imageless computer-assisted (CAS) TKA to conventional TKA. Results indicated significantly better mechanical alignment and significantly higher Knee Society Scores in favour of CAS over control groups. Femoral component alignment was significantly better with CAS, but tibial component alignment was comparable with conventional TKA. Operative time was significantly longer with CAS compared to conventional methods.