The purpose of this study was to estimate the survivorship of the varus-valgus constrained (VVC) knee implants in primary, aseptic, and septic revision total knee arthroplasty (TKA); determine functional outcomes; main modes of failure; and variables associated with increased mechanical failures. In this study, 685 consecutive cases of primary (n = 247), aseptic (n = 315), and septic revision (n = 123) TKAs with VVC implants were performed between 1999 and 2008; 533 knees (78%) had a mean follow-up of 8.2 years (range, 2-15.1). Kaplan-Meier method was used to evaluate implant survival with mechanical failure as the end point. Clinical outcomes were measured with a modified Knee Society Score (mKSS) and modified Knee Function Score (mKFS) and modes of failure were determined. Cox proportional hazards models were performed to assess for factors associated with implant failure. Ten-year survival was 88.5% (95% confidence interval [CI]: 83.9-93.5%) for primary TKAs, 75.8% (95% CI: 70.4-81.7%) for aseptic, and 54.6% (95% CI: 43.7-68.2%) for septic revisions. Improvement in pre- to postoperative mKSS and mKFS were significant in all three groups (p < 0.05). The most common mode of failure overall was infection. Mechanical modes of failures included periprosthetic fracture (45%) for primary TKA and soft tissue instability (19%) for aseptic revisions. A longer period since the last surgery in affected knee was associated with lower mechanical failures (hazards ratio of 0.55 [95% CI: 0.31-0.95], p = 0.03). VVC implant showed reliable survivorship at 10 years although careful patient selection is warranted due to the risk of infection. The main mechanical modes of failure were instability and periprosthetic fracture.