Sodium–glucose cotransporter 2 (SGLT2) inhibitors are increasingly prescribed for the treatment of patients with type 2 diabetes to reduce the risk of cardiovascular events, including heart failure (HF). The mechanisms by which SGLT2 inhibitors reduce such risk are likely to be independent of diabetes status and improvement of glycemic control. In this commentary, based on recent mediation analyses of cardiovascular outcome trials with SGLT2 inhibitors, we discuss the prognostic role of a well-known HF-related biomarker, amino-terminal pro-B-type natriuretic peptide (NT-proBNP), in patients receiving SGLT2 inhibitors. Interestingly, the NT-proBNP concentration had a relatively small impact on the SGLT2 inhibitor-associated benefit on HF events, suggesting a limited value in measuring NT-proBNP concentrations to monitor effects on cardiovascular outcomes after initiation of SGLT2 inhibitor therapy. Instead, clinical factors, such as body weight and volume status, were prognostic for cardiovascular outcomes. As shown in some biomarker studies, short-term SGLT2 inhibitor treatment significantly improved volume and HF-related health status, despite the absence of a significant change in NT-proBNP concentration. Given the early and continuous risk reduction in HF events seen in the cardiovascular outcome trials with SGLT2 inhibitors, changes in these fundamental clinical parameters after initiation of SGLT2 inhibitor therapy, independent of NT-proBNP, could be more prognostic and could represent key determinants to identify responders or non-responders to SGLT2 inhibitors for cardiovascular outcomes. Thus, this commentary highlights the clinical importance of establishing how clinicians should monitor patients initiating SGLT2 inhibitor therapy to predict the expected cardiovascular benefit. Further detailed investigations and discussion to better understand this ‘‘black box’’ are urgently warranted.