Fifty-two patients with knee osteoarthritis were randomized to receive either a conventional total knee arthroplasty (CTKA; n=26) or an imageless handheld robotic-assisted total knee arthroplasty (RATKA; n=26). The primary outcome of interest was the accuracy of the achieved hip-knee-ankle (HKA) axis alignment postoperatively. Secondary outcomes included the accuracy of femoral component varus, femoral component external rotation, femoral component flexion, and tibial component posterior slope. Outcomes were assessed at six weeks postoperatively. Overall, the results of the study revealed that the RATKA group had significantly fewer alignment outliers and more accurate component positioning. The findings suggest that the RATKA system improves implant placement accuracy in TKA.