Objective. To examine whether the consistency or persistence of knee pain, in addition to its severity, predicts incident total knee replacement (TKR). Methods. The Multicenter Osteoarthritis Study (MOST) is a longitudinal study of persons aged 50 to 79 years with symptomatic knee osteoarthritis or at high risk of disease. Subjects were queried about the presence of knee pain on most days of the previous 30 days (i.e., frequent knee pain; FKP) at 2 timepoints: a telephone screen followed by a clinic visit (median separation 4 weeks). We defined a knee as having "consistent pain" if the subject answered positively to the FKP question at both timepoints, "inconsistent pain" if FKP was positive at only one timepoint, or as "no FKP" if negative at both. We examined the association between consistent FKP and risk of TKR using multiple binomial regression with generalized estimating equations. Results. In 3026 persons (mean age 63 yrs, mean body mass index 30.4), 2979 knees (50%) had no FKP at baseline, 1279 knees (21.5%) had inconsistent FKP, and 1696 knees (28.5%) had consistent FKP. Risk of TKR over 30 months was 0.8%, 2.6%, and 8.8% for knees with no, inconsistent, and consistent FKP, respectively. Relative risks of TKR over 30 months were 1.2 (95% CI 0.6-2.3) and 2.3 (95% CI 1.2-4.4) for knees with inconsistent and consistent FKP, compared with those without FKP. This association was consistent across each level of pain severity on the Western Ontario and McMaster Universities Osteoarthritis Index. Conclusion. Consistency of frequent knee pain is associated with an increased risk of TKR independently of knee pain severity. The Journal of Rheumatology Copyright © 2011. All rights reserved.