Abstract <p>This study aimed to determine whether robotic-assisted total knee arthroplasty (RATKA) provides superior patient-reported outcomes compared to conventional manual total knee arthroplasty (MTKA), with particular emphasis on the forgotten joint score (FJS) and postoperative satisfaction levels. In this prospective longitudinal observational study, we enrolled 157 consecutive patients who underwent primary TKA between May and July 2023. The participants were assigned to either the RATKA or MTKA groups based on the surgical technique utilized. Following exclusions for periprosthetic joint infection (<i>n</i> = 2) and inadequate follow-up compliance (defined as &gt; 5 missed visits, <i>n</i> = 26), the final analytic cohort comprised 129 participants. The RATKA group (<i>n</i> = 55) demonstrated significantly higher preoperative expectations (<i>P</i> = 0.011) and longer operative duration (<i>P</i> &lt; 0.001) compared to the MTKA group (<i>n</i> = 74), while all other baseline characteristics were comparable between the groups. The mean FJS increased from preoperative values of 21.07 and 20.06 to 73.46 and 72.08 in the RATKA and MTKA groups, respectively (<i>P</i> &lt; 0.001). During follow-up from 2 to 12 months, the RATKA group consistently showed higher FJS than the MTKA group at all time points; however, the differences neither reached the minimal clinically important difference (MCID) nor attained statistical significance (<i>P</i> = 0.458). At 12 months, satisfaction was numerically higher in the RATKA group but was not statistically significant (94.5% versus 91.9%, <i>P</i> &gt; 0.05). Our findings revealed no statistically significant or clinically meaningful difference in FJS improvement between both groups; however, the RATKA group showed a numerically higher point estimate than the MTKA group.</p> Trial registration <p>ChiCTR2500107625. Registration date 15/08/2025. Retrospectively registered.</p>