Abstract Background Sodium-glucose co-transporter 2 inhibitors (SGLT2i) currently represent a pillar of heart failure (HF) treatment. However, cancer patients have not been included in landmark randomized controlled trials (RCTs), and data on safety and effectiveness in this population are lacking. Purpose Purpose of present study was to evaluate safety and effectiveness of SGLT2i in active cancer patients with HF. Methods TOSCA is a multi-centre observational trial including patients (≥ 18 years) with active cancer (diagnosis < 3 years) receiving SGLT2i for HF treatment. Patients were enrolled in two (prospective and retrospective) cohorts. The primary endpoint was safety (incidence of known and unexpected side effects). The secondary endpoint was effectiveness (ejection fraction (EF) increase, New York Heart Association (NYHA) class change, HF-related events). Exploratory endpoints included drug-drug interactions with anticancer and supportive care drugs, treatment of cancer therapy-related cardiac dysfunction (CTRCD), and changes in NT-proBNP. Results Eighty-three pts were enrolled (59 prospective - 24 retrospective). Median age was 71 (range 44-92) yrs, M/F 48/52%. The main cancer sites were breast, gastrointestinal tract, and hematological malignancies. Most cases (62%) received an active treatment (mainly chemotherapy) while on SGLT2i. Prevalent aetiology was drug-induced HF (37%) with reduced EF as the prevalent clinical presentation (39%). Prescribed agents were dapagliflozin in 60% and empagliflozin in 40% of cases on top of conventional guideline-directed medical therapy, with a median treatment duration of 3 (range 3-25) months. The incidence of urinary tract infection was 2.5%. No cases of genital infection, hypoglycemia, diabetic ketoacidosis, acute renal injury, thrombosis, and bone fractures were reported. Mean overall EF slightly increased (44.1% vs. 46.6%), and NYHA class improved (class improvement in 16%). Unplanned cardiology visits (1.2%), i.v. diuretics (1.2%), coronary angiography (5%), emergency access (2.5%), and new HF episodes (3.7%) rates were low. SGLT2i appeared effective in 31 cases of CTRCD. A relative reduction in mean NT-proBNP levels (2748 vs 1125pg/mL) was observed. No drug-drug interactions were reported. Conclusions SGLT2i appear to be safe and effective in active cancer patients with HF, with potential cardioprotective effect. No new safety warnings were recorded.