BACKGROUND: Muscle wasting predicts mortality in patients with end-stage renal disease (ESRD), but its role in the progression of chronic kidney disease (CKD) is uncertain. We studied CKD outcomes associated with low muscle mass, assessed by urinary creatinine excretion (UCr). METHODS: The NephroTest cohort included 1429 patients with CKD stages 1-4 and both measured glomerular filtration rate (mGFR) (by (51)Cr-EDTA) and estimated glomerular filtration rate (eGFR) (by CKD-Epidemiology Collaboration equation). We used cause-specific Cox models to estimate hazard ratios (HRs) for the competing risks of ESRD and death associated with gender-specific UCr quartiles. RESULTS: UCr was 13.6 +/- 3.2 mmol/24 h (0.17 +/- 0.05 mmol/kg/24 h) in men and 9.2 +/- 2.1 (0.14 +/- 0.05) in women. It was positively associated with mGFR, but not with eGFR. Over a median follow-up of 3.6 (2.1-5.8) years, 229 patients developed ESRD and 113 patients died before ESRD. Compared with patients in the highest UCr quartile, those in the lowest quartile had a higher crude HR (95% confidence interval) for pre-ESRD death: 4.3 (2.4-7.7), which was weakened, but remained statistically significant, independent of demographics, mGFR and several other factors: 2.1 (1.04-4.3). Their crude ESRD risk was not higher: HR: 0.95 (0.65-1.4), and even tended to be lower after adjusting for mGFR and log-proteinuria: HR: 0.70 (0.45-1.1). Adjustment for eGFR instead of mGFR reversed this relationship: HR: 1.7 (1.1-2.7). CONCLUSIONS: In early stage CKD, low UCr is associated with higher risk for mortality, but not for ESRD. Using creatinine-based equation to adjust for GFR may bias the relationship of UCr with ESRD risk.