Use of serum creatinine to estimate GFR may lead to underestimation of the association between self-reported frailty and kidney function. Our objectives were to evaluate the association of measured GFR (mGFR) with self-reported frailty among patients with CKD and to determine whether self-reported frailty was associated with death after adjusting for mGFR.Participants in the Modification of Diet in Renal Disease study (1989-1993) had GFR measured using iothalamate clearance (mGFR), and GFR was estimated based on the CKD-EPI creatinine (eGFRcr) and cystatin C (eGFRcys) equations. We defined self-reported frailty as three or more of: exhaustion, poor physical function, low physical activity, and low body weight. Death was ascertained through 2007 using the National Death Index and the United States Renal Data System.Eight hundred twelve MDRD participants (97 %) had complete data on self-reported frailty (16 % prevalence, N = 130) and mGFR (mean (SD) 33.1 ± 11.7 ml/min/1.73 m(2)). Higher GFR was associated with lower odds of self-reported frailty based on mGFR, (OR 0.71, 95 % CI 0.60-0.86 per 10 ml/min/1.73 m(2)), eGFRcr (OR 0.80, 95 % CI 0.67-0.94 per 10 ml/min/1.73 m(2)), and eGFRcys (OR 0.75, 95 % CI 0.62-0.90 per 10 ml/min/1.73 m(2)). Median follow-up was 17 (IQR 11-18) years, with 371 deaths. Self-reported frailty was associated with a higher risk of death (HR 1.71, 95 % CI 1.26-2.30), which was attenuated to a similar degree when mGFR (HR 1.48, 95 % CI 1.08-2.00), eGFRcr (HR 1.57, 95 % CI 1.15-2.10), or eGFRcys (HR 1.51, 95 % CI 1.10-2.10) was included as an indicator of kidney function.We found an inverse association between kidney function and self-reported frailty that was similar for mGFR, eGFR and eGFRcys. In this relatively healthy cohort of clinical trial participants with CKD, using serum creatinine to estimate GFR did not substantially alter the association of GFR with self-reported frailty or of self-reported frailty with death.