Background Myocardial fibrosis is characterized by excessive cross‐linking and deposition of collagen type I and is involved in left ventricular stiffening and left ventricular diastolic dysfunction (LVDD). We investigated whether the effect of spironolactone on LVDD in patients with heart failure with preserved ejection fraction (HFpEF) depends on its effects on collagen cross‐linking and/or deposition. Methods and results We investigated 381 HFpEF patients from the multicentre, randomized, placebo‐controlled Aldo‐DHF trial with measures of the E:e' ratio. The ratio of serum carboxy‐terminal telopeptide of collagen type I to serum matrix metalloproteinase‐1 (CITP:MMP‐1, an inverse index of myocardial collagen cross‐linking) and serum carboxy‐terminal propeptide of procollagen type I (PICP, a direct index of myocardial collagen deposition) were determined at baseline and after 1‐year treatment with spironolactone 25 mg once daily or placebo. Patients were classified by CITP:MMP‐1 and PICP tertiles at baseline. While CITP:MMP‐1 tertiles at baseline interacted (P < 0.05) with spironolactone effect on E:e', PICP tertiles did not. In fact, while spironolactone treatment did not modify E:e' in patients with lower CITP:MMP‐1 levels, this ratio was significantly reduced in the remaining spironolactone‐treated patients. In addition, PICP was unchanged in patients with lower CITP:MMP‐1 levels but was reduced in the remaining spironolactone‐treated patients. Conclusions A biochemical phenotype of high collagen cross‐linking identifies HFpEF patients resistant to the beneficial effects of spironolactone on LVDD. It is suggested that excessive collagen cross‐linking, which stabilizes collagen type I fibres, diminishes the ability of spironolactone to reduce collagen deposition in these patients.