Objective To investigate the effect of gender differences in distal femoral geometry on the clinical outcomes of total knee arthroplasty (TKA). Methods From March 2003 to October 2006, 213 cases with 294 osteoarthritic knees who had underwent primary TKA were involved in the study. The study included 87 men (118 knees) and 126 women (176 knees) who had a mean age of 64.4 years. All the cases were assessed with the knee-rating system of Knee Society, the knee range of flexion (ROF), corrected posterior offset (PCO) and anterior condylar offset (ACO) based on sex. Results Each patient had a Minimum of 2 years follow-up. The preoperative Knee Society scores and ROF, the two-year postoperative changes of stability scores and walking ability had no differences between the two groups. While greater improvements of postoperative pain scores and stair-climbing ability were seen in male. Men had better intraoperative ROF than women had ones, but there were no gender differences with regard to the two-year postoperative improvement of ROF. Corrected PCO decreased more markedly in female than in male. Corrected ACO decreased equally in female and male. The difference in the posterior condylar offset after TKA was statistically correlated with the change in pain scores, stair climbing ability and intraoperative ROF, respectively. Conclusion Male had a better outcome than female did after TKA. Femoral components which designed in the light of Caucasian anatomic characteristics couldn't match the native anatomy of distal femurs of Chinese female. Sexual dimorphism in humans and anatomic variations in various ethnic groups should be seriously considered in total knee prosthesis design.