Purpose This study compared the 5-year results of posterior cruciate ligament (PCL)-sacrificing total knee arthroplasty (TKA) with either a post and cam posterior-stabilized (PS) device, a dished, congruent condylar-stabilizing (CS) device, or a deep-dished ultra-congruent (UC) device. The hypothesis was that the clinical and radiographic outcomes would be equivalent. CS and PS participants were part of a prospective, randomized trial, and UC participants were part of a separate prospective, non-randomized protocol that was otherwise identical. A kinematic alignment surgical technique was utilized. Methods Participants were assessed preoperatively, and postoperatively at 6 weeks, 6 months, and annually for 5 years by Knee Society Score (KSS), SF-36 v2, Lower Extremity Activity Scale (LEAS), and physical and radiographic evaluation. There were 116 CS/PS participants and 69 UC participants who participated in the study. Results Tourniquet (P = .02) and operative (P = .01) times for the CS and UC groups were significantly shorter than the PS group. KSS Function scores were better for the UC group than the CS and PS groups at 6 months (P = .04) and 1 year (P = .03), and better in the UC group vs. CS at 2 years (P = .04). The KSS Pain-only score was also better in the UC compared to PS at 6 months (P = .04). There were no significant differences for the KSS Pain/Motion scores, flexion, SF-36, and LEAS scores at any time. Conclusion These data confirm the hypothesis that there are no clinically meaningful significant differences in outcomes between the three groups at a 5-year minimum follow-up, though there is a trend toward less pain and better function at earlier visits in the UC group. Level of evidence II.