BACKGROUND: Individuals with type 2 diabetes mellitus are at increased risk for heart failure (HF), particularly those with coexisting atherosclerotic cardiovascular disease and/or kidney disease. Some but not all dipeptidyl peptidase-4 inhibitors have been associated with increased HF risk. We performed secondary analyses of HF and related outcomes with the dipeptidyl peptidase-4 inhibitor linagliptin versus placebo in CARMELINA (The Cardiovascular and Renal Microvascular Outcome Study With Linagliptin), a cardiovascular outcomes trial that enrolled participants with type 2 diabetes mellitus and atherosclerotic cardiovascular disease and/or kidney disease. METHODS: Participants in 27 countries with type 2 diabetes mellitus and concomitant atherosclerotic cardiovascular disease and/or kidney disease were randomized 1:1 to receive once daily oral linagliptin 5 mg or placebo, on top of standard of care. All hospitalization for HF (hHF), cardiovascular outcomes, and deaths were prospectively captured and centrally adjudicated. In prespecified and post hoc analyses of HF and related events, Cox proportional hazards models adjusting for region and baseline history of HF were used. Recurrent hHF events were analyzed using a negative binomial model. In a subset of participants with left ventricular ejection fraction captured within the year before randomization, HF-related outcomes were assessed in subgroups stratified by left ventricular ejection fraction > or ≤50%. RESULTS: CARMELINA enrolled 6979 participants (mean age, 65.9 years; estimated glomerular filtration rate, mL/min per 1.73m; hemoglobin A1c, 8.0%; 62.9% men; diabetes mellitus duration, 14.8 years), including 1873 (26.8%) with a history of HF at baseline. Median follow-up was 2.2 years. Linagliptin versus placebo did not affect the incidence of hHF (209/3494 [6.0%] versus 226/3485 [6.5%], respectively; hazard ratio [HR], 0.90; 95% CI, 0.74-1.08), the composite of cardiovascular death/hHF (HR, 0.94; 95% CI, 0.82-1.08), or risk for recurrent hHF events (326 versus 359 events, respectively; rate ratio, 0.94; 95% CI, 0.75-1.20). There was no heterogeneity of linagliptin effects on hHF by history of HF at baseline, baseline estimated glomerular filtration rate or urine albumin-creatinine ratio, or prerandomization left ventricular ejection fraction. CONCLUSIONS: In a large, international cardiovascular outcome trial in participants with type 2 diabetes mellitus and concomitant atherosclerotic cardiovascular disease and/or kidney disease, linagliptin did not affect the risk of hHF or other selected HF-related outcomes, including among participants with and without a history of HF, across the spectrum of kidney disease, and independent of previous left ventricular ejection fraction. CLINICAL TRIAL REGISTRATION: URL: https://www.clinicaltrials.gov . Unique identifier: NCT01897532.