Objective To investigate the management of extention apparatus and cl inical results of total knee arthroplasty (TKA) of osteoarthritis. Methods Between June 2007 and June 2009, 386 patients (460 knees) with osteoarthritis received TKA, including 216 left knees and 244 right knees. There were 125 males (145 knees) and 261 females (315 knees) with an average age of 60.3 years (range, 58-85 years). The disease duration was 4-12 years (mean, 6.7 years). The X-ray films and CT scanning showed that all patients had femur-tibia joint degeneration and osteophyte formation at the edgeof joint. According to modified Burnett patellar resurfacing indication, whether or not to replace the patellar was determined, and the patellar track was determined by combining no thumb test and towel clamp traction test. Patella resurfacing was performed in 53 cases (56 knees, resurfacing group), no patella resurfacing in 333 cases (404 knees, non-resurfacing group), and lateral retinacular releasing 68 cases (72 knees). The postoperative imaging and knee society score (KSS), patellofemoral compl ications were analyzed. Results All wounds healed by first intention. All the patients were followed up 1-3 years (mean, 26 months). Deep infection occurred in 3 cases (1 case of resurfacing group and 2 cases of non-resurfacing group) at 3-7 months and they were cured after two-stage reversion. Patellofemoral compl ications occurred in 2 cases of resurfacing group (2/56, 3.57%) and in 9 cases of non-resurfacing group (9/404, 2.23%), showing no significant difference (χ2=0.38, P=0.54). There were significant differences in KSS and visual analogue scale (VAS) at 2 and 3 year after operation when compared with before operation (P lt; 0.05). No significant difference in KSS and VAS of resurfacing group and non-resurfacing group were observed at 3 years after operation (P gt; 0.05). Conclusion It is benefit for selective patellar resurfacing and the patellar tract improvement to select patellar resurfacing indication according to the multi-factor assessment and to determine the patellar tract by combining no thumb test and towel clamp traction test, which can reduce the patellofemoral compl ication rate after TKA.