AIMS: Urinary kidney injury molecule-1 (KIM-1) is a marker of tubular damage and associated with worse outcome in heart failure (HF). Plasma KIM-1 has not been described in HF. METHODS AND RESULTS: In a renal mechanistic cohort of 120 chronic HF patients, we established the association between plasma KIM-1, renal invasive haemodynamic parameters {renal blood flow ([(131) I]hippuran clearance) and measured glomerular filtration rate (GFR; [(125) I]iothalamate)} and urinary tubular damage markers. The association between plasma KIM-1, plasma creatinine, and clinical outcome was further explored in a cohort of 2033 acute HF patients. Median plasma KIM-1 was 171.5 pg/mL (122.8-325.7) in chronic (n = 99) and 295.1 pg/mL (182.2-484.2) in acute HF (n = 1588). In chronic HF, plasma KIM-1 was associated with GFR (P < 0.001), creatinine, and cystatin C. Plasma KIM-1 was associated with urinary N-acetyl-β-d-glucosaminidase (NAG), but not with other urinary tubular damage markers. Log plasma KIM-1 predicted adverse clinical outcome after adjustment for age, gender, and GFR [hazard ratio (HR) 1.94, 95% confidence interval (CI) 1.07-3.53, P = 0.030]. Statistical significance was lost after correction for NT-proBNP (HR 1.61, 95% CI 0.81-3.20, P = 0.175). In acute HF, higher plasma KIM-1 levels were associated with higher creatinine, lower albumin, and presence of diabetes. Log plasma KIM-1 predicted 60-day HF rehospitalization (HR 1.27, 95% CI 1.03-1.55, P = 0.024), but not 180-day mortality or 60-day death or renal or cardiovascular rehospitalization. CONCLUSIONS: Plasma KIM-1 is associated with glomerular filtration and urinary NAG, but not with other urinary tubular damage markers. Plasma KIM-1 does not predict outcome in chronic HF after correction for NT-proBNP. In acute HF, plasma KIM-1 predicts HF rehospitalization in multivariable analysis.