Background and Purpose: Inflammation plays a crucial role in the development and progression of numerous acute and chronic diseases such as myocardial infarction (MI) and heart failure. Targeting ADAM proteases, particularly ADAM10, holds promise for fine-tuning inflammatory responses. However, developing selective inhibitors has proven challenging because of cross-inhibition of other metalloproteases. Protein-based inhibitors, such as the synthetic ADAM10 prodomain, offer potential for improved selectivity and effectiveness. Experimental Approach: A novel ADAM10 inhibitor, VTH144, was designed by modifying the ADAM10 prodomain and evaluated empirically for selectivity, efficacy and toxicity both in vitro and in vivo and by using machine learning–based modelling. Key Results: AlphaFold3 predictions confirm that the ADAM10 prodomain is highly selective for its protease and identify VTH144 as a unique ADAM10-inhibiting peptide. Fluorescence resonance energy transfer (FRET) assays verified VTH144 as highly selective for ADAM10, sparing other metalloproteases. In a MI mouse model, we show that VTH144 applied twice, 4 and 48 h after ligation of the left ventricular descending artery (LAD), is highly effective in preserving cardiac function by reducing scar size. VTH144 significantly attenuates ADAM10-mediated CX₃CL1 shedding, subsequent neutrophil infiltration into the heart and IL-1β-driven systemic inflammation. Importantly, safety pharmacology in naïve mice revealed that ADAM10 inhibition with VTH144 lacks toxic short-term effects in cardiac, pulmonary, hepatic or renal systems. Conclusions and Implications: Our study highlights the potential of the selective ADAM10 inhibitor VTH144 as a promising therapeutic candidate for improving cardiac function and reducing the risk of heart failure following myocardial infarction.