Abstract Background Coronary microvascular dysfunction is an heterogenous entity with structural dysfunction present in about 60% of cases. Although Coronary Flow Reserve (CFR) is commonly used for diagnosis, this index is nonspecific to the microvasculature. The effectiveness of the recently introduced Microvascular Resistance Reserve (MRR) and absolute microvascular resistance (Rµ) in identifying microvascular dysfunction endotypes remains uncertain. Objective To evaluate whether MRR and hyperemic Rµ assessment are effective in dichotomizing functional and structural microvascular dysfunction as compared with Rµ and coronary blood flow (Q) monitoring during exercise. Methods Rµ and MRR were first assessed using continuous thermodilution with saline infusion at 10 mL/min (rest) and 20 mL/min (hyperemia) in the left anterior descending artery of consecutive patients with Angina and Non-Obstructive Coronary Arteries. An MRR<2.3 was considered the cutoff for coronary microvascular dysfunction. Q and Rµ were assessed during subsequent stress testing with an incremental workload of 30 watts every two minutes, using a dedicated supine cycling ergometer. Results Coronary microvascular dysfunction was observed in 36.6% of the participants with suspected Angina with Non-Obstructive Coronary Artery disease (11 out of 30). Median hyperemic Rµ was 561 [381; 684] Wood Unit (WU) among patients with microvascular dysfunction. Participants with structural dysfunction (MRR<2.3 and hyperemic Rµ>561WU), had a resting Rµ similar to participants without CMD (1286±410 versus 1231±291 WU) but exhibited both impaired reduction of Rµ (139 [64; 288] versus 191 [74; 317] WU per 30 watts) and increase of Q during exercise (20 [1; 37.5] versus 52 [15.5; 92.0] mL/min per 30 watts). Patients with functional dysfunction (MRR<2.3 and hyperemic Rµ<561WU) had lower resting Rµ (672±219 WU) and higher resting Q (152±39 versus 84±24 mL/min) with a similar increase of Q during exercise (55 [30.5; 103.0] than participants without CMD. Mean N-terminal pro B-type natriuretic peptide level was higher in the structural group (425±563 ng/L) as compared to the functional and normal groups (86±10 ng/L and 50±80 ng/L respectively). Conclusion The use of MRR combined to hyperemic Rµ, derived from continuous intracoronary thermodilution, enables the assessment of coronary microvascular dysfunction with an accurate distinction between functional and structural endotypes.Microvascular hemodynamic adaptationNTpBNP level