BACKGROUND: Total knee replacement (TKR) is clinically and cost-effective. The surgical approach employed influences the outcome, however there is little generalisable and robust evidence to guide practice. We compared outcomes between the common primary TKR surgical approaches. METHODS: 875,166 primary TKRs captured in the National Joint Registry, linked to hospital inpatient, mortality and patient reported outcome measures (PROMs) data, with up to 15.75 years follow-up were analysed. There were 10 surgical approach groups: medial parapatellar, midvastus, subvastus, lateral parapatellar, 'other' and their minimally invasive versions. Survival methods were used to compare revision rates and 45-day mortality. Groups were compared using Cox proportional hazards regression and Flexible Parametric Survival Modelling (FPM). Confounders included age at surgery, sex, risk group (indications additional to osteoarthritis), American Society of Anesthesiologists grade, TKR fixation, year of primary, body mass index, and for mortality, deprivation and Charlson comorbidity subgroups. PROMs were analysed with regression modelling or non-parametric methods. RESULTS: The conventional midvastus approach was associated with lower revision rates (Hazard Rate Ratio (HRR) 0.80 (95% CI 0.71-0.91) P = 0.001) and the lateral parapatellar with higher revision rates (HRR 1.35 (95% CI 1.12-1.63) P = 0.002) compared to the conventional medial parapatellar approach. Mortality rates were similar between approaches. PROMs showed statistically significant, but not clinically important, differences. CONCLUSIONS: There is little difference in PROMs between the various surgical approaches in TKR with all resulting in good outcomes. However, the conventional midvastus approach (used in 3% of cases) was associated with a 20% reduced risk of revision surgery compared to the most commonly used knee approach (the conventional medial parapatellar: used in 91.9% of cases). This data supports the use of the midvastus approach and thus surgeons should consider utilising this approach more frequently. Minimally invasive approaches did not appear to convey any clinical advantage in this study over conventional approaches for primary TKR.