Objective To evaluate the usefulness of minimal incision technique in total knee arthroplasty (TKA) by comparing the early compl ications after minimal incision TKA and those after traditional incision. Methods From May 2004 to July 2005, 38 patients (46 knees) underwent TKA using the minimal incision technique (minimal incision group), and 43 patients (54 knees) underwent TKA using the traditional incision technique at the same period (traditional incision group). The inimal incision group included 12 male patients (12 knees) and 26 female patients (34 knees), and their ages ranged from 52 to 76 years. Twenty-four patients (28 knees) had osteoarthritis and 14 patients (18 knees) had rheumatic arthritis. The varus deformity of the knee was found in 30 patients (34 knees) and valgus deformity was found in 8 patients (12 knees). TheAmerican Knee Society Score (AKSS) score was 37.5 ± 12.6, and the disease course was (7.5 ± 2.3) years. The raditional incision group included 15 male patients (19 knees) and 28 female patients (35 knees), and their ages ranged from 55 to 82 years. Thirtytwo patients (37 knees) had osteoarthritis and 11 patients (17 knees) had rheumatic arthritis. Varus deformity of the knee was found in 34 patients (41 knees) and valgus deformity was found in 9 patients (13 knees). The AKSS score was 31.1 ± 10.2, and the disease course was (10.1 ± 4.2) years. There were no statistically significant differences in the general data between two groups (P gt; 0.05). Results The incision length, the operation time and the drainage flow were (12.6 ± 1.2) cm, (95 ± 15) minutes and (650.1 ± 10.0) mL in the minimal incision group and (18.7 ± 2.3) cm, (63 ± 11) minutes and (300.0 ± 20.0) mL in the traditional incision group; showing statistically significant differences between two groups (P lt; 0.05). In the minimal incision group, 4 patients (4 knees) developed infections at the operated knees, including 2 early infection and 2 late infection, which were all cured by corresponding treatment. Deep vein thrombosis occurred in 1 patient on the third day after operation and was managed successfully by thrombolytic therapy. Cutaneous necrosis was found in 2 patients on the seventh and ninth postoperative day separately, which healed uneventfully after intensive local treatment. On the twelfth postoperative month, 1 patient had femoral fractured at the site of supracondylar region after a careless fall, but the prosthesis was stable. The fracture was fixed by a plate and healed uneventfully. In the traditional incision group, only 1 patient (1 knee) developed early infection at the operated knee on the tenth postoperative day, which was managed by corresponding treatment. And there were no periprosthetic fracture, cutaneous necrosis or deep vein thrombosis. The patients were followed up for (3.7 ± 0.4) years in the minimalincision group and (3.9 ± 0.6) years in the traditional incision group. At the latest follow-up, the AKSS scores were 78.2 ± 6.7 in the minimal incision group and 81.2 ± 7.3 in the traditional incision group, showing statistically significant ifferences (P lt; 0.05) when compared with those before operation and no statistically significant difference between two groups (P gt; 0.05). Conclusion Minimally invasive TKA has relatively higher compl ication rate than traditional incision. Strict patient inclusion criteria, competent surgery skill, proper instrument and intensive perioperative management are essential to success.