Introduction Tibiofemoral dislocation is the most serious form of instability following total knee arthroplasty (TKA). It is little reported in the literature, despite severity comparable to that in the native knee. The present systematic review and meta-analysis aimed to identify risk factors and treatment strategies. Methods The Cochrane, Medline (via PubMed), Google, and PROSPERO data-bases were searched in January 2018 following the PRISMA meta-analysis guidelines. All articles referring to tibiofemoral dislocation following primary TKA were included for analysis and extraction of individual data. Study data comprised age, gender, comorbidities, primary and revision implant design, aetiology, and treatment strategy. Results Individual data for 57 patients (23 studies) were analyzed. A total of 62.0% of primary implants were posterior-stabilized (49 patients) and 30.4% posterior-cruciate-retaining (24 patients). Obesity was the most frequent comorbidity (39.2%; 31 patients), followed by severe pre-operative deformity (31.6%; 25 patients). Aetiologies of dislocation were mainly related to comorbidity (15 studies; 65.2%) or intra-operative iatrogenic destabilization (14 studies; 60.9%). Non-operative treatment (splint) after dislocation was associated with high rates of recurrence (39.1%) but significantly fewer complications (p = 0.033). Implant revision surgery (45 patients; 80.4%) usually involved higher-constraint models (31 patients; 70.8%). Conclusions Improved implant design has reduced the rate of tibiofemoral dislocation, although this complication remains serious in both the short and the long term. The present review identified patient- and surgeon-related risk factors. Awareness of the former, which are identifiable pre-operatively, and of the most frequent technical errors is critical.