BACKGROUND: Diuretics are the mainstay of treatment for congestion but concerns exist that they adversely affect prognosis. We explored whether the relationship between loop diuretic use and outcome is explained by the underlying severity of congestion amongst patients referred with suspected heart failure. METHOD AND RESULTS: Of 1190 patients, 712 had a left ventricular ejection fraction (LVEF) ≤50 %, 267 had LVEF >50 % with raised plasma NTproBNP (>400 ng/L) and 211 had LVEF >50 % with NTproBNP ≤400 ng/L; respectively, 72 %, 68 % and 37 % of these groups were treated with loop diuretics including 28 %, 29 % and 10 % in doses ≥80 mg furosemide equivalent/day. Compared to patients with cardiac dysfunction (either LVEF ≤50 % or NT-proBNP >400 ng/L) but not taking a loop diuretic, those taking a loop diuretic were older and had more clinical evidence of congestion, renal dysfunction, anaemia and hyponatraemia. During a median follow-up of 934 (IQR: 513-1425) days, 450 patients were hospitalized for HF or died. Patients prescribed loop diuretics had a worse outcome. However, in multi-variable models, clinical, echocardiographic (inferior vena cava diameter), and biochemical (NTproBNP) measures of congestion were strongly associated with an adverse outcome but not the use, or dose, of loop diuretics. CONCLUSIONS: Prescription of loop diuretics identifies patients with more advanced features of heart failure and congestion, which may account for their worse prognosis. Further research is needed to clarify the relationship between loop diuretic agents and outcome; imaging and biochemical measures of congestion might be better guides to diuretic dose than symptoms or clinical signs.