The purpose of this study is to determine if a fellow's involvement, as well as duration of a fellow's training, impacts complications, outcomes, and survivorship in primary total knee arthroplasty (TKA). A retrospective review identified 2,790 consecutive patients (3,530 knees) who underwent primary TKA between 2003 and 2008. A 2-year minimum follow-up was available on 2,785 knees (2,195 patients). Operative data, clinical outcomes, complications, and survivorship were compared between cases with and without a fellow involved as well as comparing cases in the first quarter (Q1) of the academic year to the last three quarters (Q2–4). Mean follow-up was 9.7 years. Fellows were involved in 1,434 (41%) surgeries. Fellow cases had significantly longer tourniquet times (59.5 vs. 49 minutes, p < 0.001) and operative times (82.4 vs. 70.8 minutes, p < 0.001). Overall, there was no difference in clinical, functional, or pain outcomes between attending and fellow cases. Fellow cases during Q1 had significantly worse pain scores (p = 0.009) and clinical scores (p < 0.001). Revision rate, infection rates, and survivorship were not significantly different between fellow and attendings or during Q1 of fellowship compared with attendings. Primary TKA survivorship and complications were not affected by fellow involvement or academic quarter. An “August Effect” may be suggested for clinical and pain outcomes in the first 3 months of fellowship.