Introduction: To reduce the size of the surgical incision, modular mini-keel tibial components have been developed with or without extensions for the Nexgen (TM) MIS Tibial Component. Although a smaller component could theoretically result in defective fixation, this has never been evaluated in a large comparative series. Thus, we performed the following case control study to: (1) evaluate intermediate-term survival of a modular ``mini-keel'' tibial component compared to a reference standard keel component from the same line of products (Nexgen LPS-Flex Tibial Component, Zimmer); (2) to identify any eventual associated factors if the frequency of loosening was increased. Hypothesis: The rate of revision for aseptic tibial loosening is comparable for both components. Materials and methods: This comparative, retrospective, single center series of 459 consecutive total knee arthroplasties (TKA) was performed between 2007 and 2010: with 212 modular ``mini-keel'' (MK) tibial components and 247 ``standard'' (S) components. Survival, rate of revision for aseptic tibial loosening and identification of a radiolucent line were analyzed at the final follow-up. Results: After a median follow-up of 5years, the rate of revision for tibial aseptic loosing was significantly higher in the MK group with 12 cases (5.7%) and 4 cases in the S group (1.6%) (P=0.036). The use of the MK component appears to be a prognostic factor for surgical revision (hazard ratio=3.86 (1.23-11.88), P=0.02) but not for the development of a radiolucent line (HR=1.75 (0.9-3.4), P=0.097). The mean delay before revision was 38months (8-64) in the MK group and 15.2months (8-22) in the S group (P=0.006). Individual factors, such as gender, body mass index (BMI) and pre- or postoperative alignment were not prognostic factors for revision or radiolucent lines. Conclusion: The modular ``mini-keel'' tibial component was associated with a greater risk of revision for tibial component loosening. (C) 2016 Elsevier Masson SAS. All rights reserved.