Rotating-platform knee implants have successively undergone modifications to improve postoperative flexion. The cruciate-sacrificing Low Contact Stress (LCS) implant (DePuy Orthopaedics, Inc, Warsaw, Indiana) was modified into the cruciate-substituting PFC Sigma RP (SRP) implant and further into the PFC Sigma RPF (SRPF) implant (DePuy Orthopaedics, Inc). The goal of this study was to determine whether these modifications improved postoperative flexion. Postoperative flexion at 2 years was compared against preoperative flexion with regard to the general demographics of each group.Statistical analysis showed that the pre- to postoperative flexion changes achieved by the SRP (14.6°) and the SRPF (2.9°) were better (P<.001) than that achieved by the LCS (-10.3°); however, between the SRP (14.6°) and the SRPF (2.9°), the change was statistically insignificant (P=.045). In subgroups with preoperative flexion less than 125°, postoperative flexion achieved was 100.1° with the LCS, 119.8° with the SRP, and 121.3° with the SRPF. The difference between the SRP and SRPF and the LCS was statistically significant (P<.001), but between the SRP and the SRPF was statistically insignificant (P=.621). In subgroups with preoperative flexion 125° or more, postoperative flexion was 125° with the LCS, 132° with the SRP, and 130° with the SRPF, with no significant difference between groups (P=.416). Both cruciate-substituting designs produced better postoperative flexion than the cruciate-sacrificing design. The SRP, despite less preoperative flexion (P=.004), achieved statistically better postoperative flexion than the LCS (P<.001). In subgroups with comparable preoperative flexion, no statistical difference in postoperative flexion was achieved by the SRP and the SRPF.