We hypothesized that navigation can help provide a well-balanced knee, through real-time feedback of alignment accuracies and gap sizes in flexion and extension. The purpose of this study was to evaluate in vivo stabilities of mediolateral laxity in full extension and anteroposterior laxities in 90° of flexion after navigation-assisted total knee arthroplasty, and to determine the nature of the correlations between these and range of motion (ROM). Forty-two total knee arthroplasties performed using a navigation system with a minimum two-year follow-up were included. The following were measured at final follow-ups; mediolateral laxities at extension and anteroposterior laxities at 90 degrees of flexion (using stress radiographs and a Telos arthrometer), modified HSS scores (excluding laxity and range of motion), and range of motion (ROM). At final follow-up the mean modified HSS score was 82% of total points and mean postoperative ROM was 128.1 ± 10.4°. Mean medial laxity was 3.5 ± 1.4°, mean lateral laxity 4.4 ± 2.2°, and mean anteroposterior laxity 7.1 ± 4.1 mm. We found no significant correlation between mediolateral laxity and postoperative ROM. However, a significant correlation was found between postoperative ROM and anteroposterior laxity. In the present study, the use of a navigation system in total knee arthroplasty was found to improve in vivo stability and produce promising short-term clinical results.