Subclinical diabetic cardiomyopathy (DCM) is frequent in asymptomatic subjects with type 2 diabetes (T2DM). We sought the response of functional and fibrosis markers to therapy in a trial of aldosterone antagonism for treatment of DCM.Biochemical, anthropometric, and echocardiographic data were measured in 225 subjects with T2DM. Myocardial function was evaluated with standard echocardiography and myocardial deformation; ischaemia was excluded by exercise echocardiography. Calibrated integrated backscatter and post-contrast T1 mapping from cardiac magnetic resonance imaging were used to assess myocardial structure. Amino-terminal propeptides of pro-collagen type I (PINP) and III (PIIINP), the carboxy-terminal propeptide of pro-collagen type I (PICP) and transforming growth factor beta-1 were measured from peripheral blood or urine to assess myocardial collagen turnover.Diastolic dysfunction was identified in 81 individuals, of whom 49 (25 male, age 60 ± 10 years) were randomized to spironolactone 25 mg/day or placebo therapy for 6 months. Groups were well-matched at baseline. Spironolactone therapy was associated with improvements in diastolic filling profile (Δpeak E wave velocity -4 ± 15 vs. 9 ± 10 ms, P = 0.001; ΔE/A ratio -0.1 ± 0.3 vs. 0.2 ± 0.2, P < 0.001) and cIB values (-21.2 ± 4.5 dB vs. -18.0 ± 5.2 dB, P = 0.026; ΔcIB -5.1 ± 6.8 vs. -1.3 ± 5.2, P = 0.030). ΔcIB was independently associated with spironolactone therapy (β = 0.320, P = 0.026) but not Δblood pressure. With intervention, pro-collagen biomarkers (ΔPINP P = 0.92, ΔPICP P = 0.25, ΔPIIINP P = 0.52, and ΔTGF-β1 P = 0.71) and T1 values (P = 0.54) remained similar between groups.Spironolactone-induced changes in myocardial structure and diastolic properties in DCM are small, and are unassociated with changes in collagen biomarkers or T1 values.