Myocardial fibrosis and microvascular dysfunction are key determinants of outcome in heart failure (HF); we examined their relationship in patients with HF. Our study included 61 consecutive patients with HF but without coronary stenosis. All underwent gadolinium-enhanced cardiac magnetic resonance to evaluate late gadolinium enhancement (LGE) and an acetylcholine (ACh) provocation test to evaluate microvascular dysfunction. During the ACh provocation test, we sampled blood simultaneously from the coronary sinus and aortic root to compare lactate concentrations. We quantified coronary blood flow volume using an intracoronary Doppler-tipped guidewire. We detected LGE in 34 patients (LGE-positive); 27 were LGE-negative. Coronary blood flow volume increased significantly after the ACh provocation test only in LGE-negative patients (before vs. after ACh, 47.5 ± 36.8 vs. 69.2 ± 48.0 ml/min, respectively; p = 0.004). The myocardial lactate extraction ratio (LER) significantly decreased after the ACh test in both groups (LGE-negative, p = 0.001; LGE-positive, p < 0.001), significantly more so in the LGE-positive group (p = 0.017). Multivariate logistic regression analysis showed that a post-ACh LER < 0 (indicating myocardial lactate production) was a significant predictor of LGE-positivity (odds ratio 4.54; 95 % confidence interval 1.38–14.93; p = 0.013). In the LGE-positive group, an LGE volume greater than the median significantly predicted a post-ACh LER of <0 (p = 0.042; odds ratio 6.02; 95 % confidence interval 1.07–33.86). ACh-provoked coronary vasomotor abnormality is closely correlated with myocardial fibrosis in patients with HF but without organic coronary stenosis. Coronary vasomotor abnormalities in fibrotic myocardium may worsen HF.