OBJECTIVES: The non-selective mineralocorticoid receptor antagonist spironolactone is an established treatment for heart failure (HF). However, some evidence suggests that it may have a deleterious effect on glucose homeostasis. The objective of this study was to assess whether spironolactone may increase the risk of developing diabetes in a large cohort of HF patients. METHODS:  We studied a population-based cohort of patients hospitalized with a primary diagnosis of HF using two administrative databases: the Quebec government administrative database of hospital discharges (MED-ECHO) and the Quebec medical services and prescription claims database (RAMQ) from January 1995 to December 2009. Patients were categorized as new users of spironolactone and non-users. The primary outcome was the new-onset diabetes. RESULTS:  The cohort consisted of 5,773 patients hospitalized with a primary diagnosis of HF, of which 873 were new users of spironolactone. The incidence of new-onset of diabetes was greater in spironolactone users (6.3 per 100 person-years) than in non-users (4.1 per 100 person-years). This increase was significant in both the crude, unadjusted model, with a hazard ratio (HR) of 1.31; 95% CI: 1.08-1.58; p = 0.005, and in an adjusted Cox proportional hazard model (HR: 1.25; 95% CI: 1.03-1.52; p = 0.0259). A propensity-matched cohort analysis revealed consistent results. CONCLUSIONS: HF patients treated with spironolactone in the community appear to have modest increased risk of new-onset of diabetes compared to non-users. Further investigations using large randomized controlled trials will be necessary to confirm the results from this population-based observational study.