To determine whether racial disparities in total knee replacement (TKR) failure are explained by poverty. Black and white New York state residents, enrolled in a prospective single-institution TKR registry January 1, 2008 to February 6, 2012, who underwent primary unilateral TKR (n = 4,062) were linked to the New York Statewide Planning and Research Cooperative System database (January 1, 2008 to December 31, 2014) to capture revisions performed at outside institutions. Patients were linked by geocoded addresses to residential census tracts. Multivariable Cox regression was used to assess predictors of TKR revision. Multivariable logistic regression was used to analyze predictors of TKR failure, defined as TKR revision in New York state ≤2 years after surgery, or as Hospital for Special Surgery (HSS) TKR quality of life score "not improved" or "worsened" 2 years after surgery. The mean ± SD age was 68.4 ± 10 years, 64% of patients were female, 8% lived in census tracts with >20% of the population under the poverty line, and 9% were black. Median follow-up time was 5.3 years. A total of 3% of patients (122 of 4,062) required revision a median 454 days (interquartile range 215-829) after surgery. TKR revision risk was higher in blacks than whites, with a hazard ratio of 1.69 (95% confidence interval 1.01-2.81), but in multivariable analysis, only younger age, male sex, and constrained prosthesis were predictors of TKR revision. TKR failure occurred in 200 of 2,832 cases (7%) with 2-year surveys. Risk factors for TKR failure were non-osteoarthritis TKR indication, low surgeon volume, and low HSS Expectations Survey score, but not black race. Community poverty was not associated with TKR revision or failure. There was a trend toward higher TKR revision risk in blacks, but poverty did not modify the relationship between race and TKR revision or failure.