To evaluate the clinical outcome in 108 total knee arthroplasty (TKA) patients with a medial-lateral balanced versus unbalanced gap after a mean follow-up of 34 months. 64 women and 44 men (mean age, 69.5 years) underwent computer-assisted TKA for osteoarthritis using a cemented fixed-bearing cruciate-retaining prosthesis. The medial-lateral gap difference (measured with the prosthesis in situ and the patella reduced) was balanced (≤2 mm) in 81 patients and unbalanced (>2 mm) in 27 patients. After a mean follow-up of 34 months, patients were assessed using the Western Ontario and McMaster Universities Arthritis Index (WOMAC) questionnaire for pain, stiffness, and physical function. Scores were normalised to 0% (worst) to 100% (best). The balanced and unbalanced gap groups were comparable in terms of age, severity of osteoarthritis, and proportion of pre- and post-operative mechanical alignment. Compared with the balanced gap group, the unbalanced gap group had a larger medial-lateral extension gap difference (0.75±0.57 vs. 2.02±1.15 mm, p=0.001) and medial-lateral flexion gap difference (0.79±0.63 vs. 2.98±2.13 mm, p=0.001) and lower normalised total WOMAC score (84.9±18 vs. 74.8±20.8, p=0.017). WOMAC score is better in TKAs with a medial-lateral balanced (<2 mm) gap.