Two surgical strategies are possible in total knee arthroplasty (TKA): a measured resection technique, in which bone landmarks are used to guide resections equal to the distal and posterior thickness of the femoral component, or a gap-balancing approach, in which equal collateral ligament tension in flexion and extension is sought before and as a guide to final bone cuts. In this study performed with computer assisted system, we compared the 2 different methods in 126 patients followed prospectively in order to analyze the effect of both the techniques on joint-line (JL) maintenance, axial limb restoration and components position. The gap technique showed a statistical increase in the post-operative value when compared with the measured resection technique, (P = 0.008). When comparing the two groups regarding to the pre-operative deformity, we have found a statistical difference (P = 0.001) in case of moderate pre-operative deformity (less than 10 degrees), and the measured resection technique showed a slight superiority in preserving a joint line more faithful to the pre-operative. We found an ideal alignment for the mechanical axis (180 degrees ± 3 degrees) (95% of cases). In six cases (5%), the mean post-operative value exceeded (varus or valgus) the ideal value by more than 3 degrees. In the frontal plane, a good alignment was observed for both femoral and tibial components without a significant difference between the two techniques. In the sagittal plane was found more alignment variability due to the different implants used and their ideal starting slope, from 7 degrees to 3 degrees. Finally, the surgeon can use the approach with which he has more confidence; however, as the measured resection technique causes less reduction in the post-operative joint-line position, in case of shortening of patellar tendon or patella infera, this technique is preferable.