We performed a retrospective clinical and radiographic evaluation of 83 nonconsecutive patients operated in our institute between February 1996 and March 2003 with a mean follow-up of 60 months to assess the efficiency of unicompartmental knee replacement (UKR) performed with a minimally invasive technique. The aim of this study was to correlate the clinical outcome with the pre- and post-operative alignment and with implant positioning on coronal and sagittal plane. Eighty-three nonconsecutive patients (60 males, 23 females) underwent cemented UKR (De Puy Preservation Uni with all-poly tibial component), for both medial OA (80 patients) and AVN of the medial femoral condyle (3 patients). All patients were available at final follow-up evaluation, and they all presented an evident varus alignment at pre-operative clinical and radiographic evaluation. At radiographic measurement, we considered a knee with femoro-tibial angle (FTA) > 175 degrees as varus knee, 170 degrees < FTA < 175 degrees as normal knee and a knee with a FTA < 170 degrees as a valgus knee. Moreover, we considered a tibial plateau angle (TPA) > 90 degrees for valgus knee and a TPA < 90 degrees for varus knee. According to Hospital for Special Surgery (HSS) scoring system, at a mean follow-up of 60 months, 61 (74%) cases were excellent (100-85 points), 15 (18%) cases were good (84-70 points) and 7 cases (8%) had fair results (<70 points). In our series, patients with an excellent clinical result presented a mean varus deformity of 7.2 degrees (3.6 degrees-10.8 degrees) pre-operatively. According to literature, we demonstrated that a small amount of undercorrection with a small amount of residual varus deformity of 3 degrees-5 degrees is the goal to be reached in order to avoid both rapid degeneration of the nonreplaced compartment and the premature loosening of the replaced compartment. We performed a mean axial correction of 5 degrees (SD 3.9 degrees), leaving a mean axial varus deformity of 2.2 degrees in the excellent group. In our series, the group with excellent results also showed a post-operative PTS of 7 degrees (2.4 degrees-11.6 degrees), while mean pre-operative PTS was 6.5 degrees (2.7 degrees-10.3 degrees). In this study, results have shown that minimally invasive UKR producing a small amount of varus undercorrection in selected patients with medial tibio-femoral osteoarthritis or moderate avascular necrosis of the medial femoral condyle provides excellent clinical and functional results. Overcorrection of varus malalignment with a UKR may produce both rapid degeneration of the lateral tibio-femoral compartment and the early failure of the replaced compartment.