Total knee arthroplasty implant designs with larger extensor moment arms theoretically should generate lower extensor forces for the same externally applied loads. This study measured knee kinematics, quadriceps forces, and patellofemoral forces under conditions of dynamic knee extension under load in two knee designs with differing quadriceps moment arms. Six human cadaver knees were tested both before implantation and after sequential implantation with two posterior cruciate retaining designs. The extensor moment arm of the LMA (long extensor moment arm design, Scorpio, Howmedica Osteonics, Rutherford, NJ) was approximately 1 cm longer than that of the Control design (7000, Howmedica Osteonics). Quadriceps tension was measured during dynamic closed kinetic chain knee extension. Patellar compressive and shear forces were also recorded using a patellar component instrumented with a custom triaxial load transducer. Knee kinematics were monitored using a three-dimensional electromagnetic tracking device. Both designs produced similar patterns of femoral rollback and tibial rotation. Quadriceps tension was lower in the LMA design compared with the Control design. Patellofemoral compressive forces were also significantly reduced in the LMA design when compared with Control (8-18% lower at angles greater than 50 degrees flexion). The design with the longer extensor moment arm required less quadriceps force to extend the knee under load and reduced patellofemoral compressive forces. Reduced quadriceps forces may facilitate postoperative rehabilitation and activities such as stair climbing. Reduction in patellofemoral forces could reduce patellar complications such as anterior knee pain, component wear, and loosening.