Aims The aim of this study was to evaluate whether biomarkers reflecting pathophysiological pathways are different between heart failure with preserved (HFpEF) and reduced ejection fraction (HFrEF) and whether the prognostic value of biomarkers is different in HFpEF vs. HFrEF. Methods and results A total of 458 HFrEF (LVEF ≤40%) and 112 HFpEF (LVEF ≥50%) patients aged ≥60 years with NYHA class ≥II from TIME‐CHF were included. Endpoints are 18‐month overall and HF hospitalization‐free survival. After correction for baseline characteristics that differed between the HF types, i.e. age, gender, body mass index, systolic blood pressure, cause of HF, and AF, HFpEF patients exhibited higher soluble interleukin 1 receptor‐like 1 ST2; 37.6 (28.5–54.7) vs. 35.7 (25.6–52.2), P = 0.02, high sensitivity C‐reactive protein (hsCRP; 8.54 (3.39–25.86) vs. 6.66 (2.42–15.39), P = 0.01), and cystatin‐C 1.94 (1.57–2.37) vs. 1.75 (1.39–2.12), P = 0.01. In contrast, HFrEF patients exhibited higher NT‐proBNP 2142 (1473–4294) vs. 4202 (2239–7411), P < 0.001, high sensitivity troponin T hsTnT; 27.7 (16.8–48.0) vs. 32.4 (19.2–59.0), P = 0.03, and haemoglobin 124 (110–135) vs. 134 (122–145), P < 0.001. In addition to these clinical characteristics, NT‐proBNP, haemoglobin, cystatin‐C, hsTnT, and ST2 improved the area under the curve from 0.86 (0.82–0.89) to 0.91 (0.87–0.94; P < 0.001) for discriminating HFpEF from HFrEF. There were no significant interactions between HFpEF and HFrEF when considering the prognostic value of the investigated biomarkers (P > 0.10 for both endpoints), except for cystatin‐C which had less prognostic impact in HFpEF (P < 0.01). Conclusion Biomarker levels suggest a different amount of activation of several pathophysiological pathways between HFpEF and HFrEF. No important differences in the prognostic value of biomarkers in HFpEF vs. HFrEF were found except for cystatin‐C, and for NT‐proBNP in the NT‐proBNP‐guided study arm only, both of which had less prognostic value in HFpEF. Trial registration ISRCTN43596477