Background The prognostic benefits of beta-blockers (BB) in patients with systolic heart failure (SHF) are known but despite this, in patients with diabetes they are underutilized. The aim of this study was to assess the effect of beta-blockers (BB) on glycaemic control in patients with Type 2 Diabetes (T2DM) and systolic heart failure (SHF) stratified to beta-1 selective (Bisoprolol) vs. nonselective BB (Carvedilol). Methods This observational, cohort study was conducted in patients with T2DM and SHF attending an Australian tertiary teaching hospital's heart failure services. The primary endpoint was glycaemic control measured by glycosylated haemoglobin (HbA1c) at initiation and top dose of BB. Secondary endpoints included microalbuminuria, changes in lipid profile and estimated glomerular filtration rate (eGFR). Results 125 patients were assessed. Both groups were well matched for gender, NYHA class and use of guideline validated heart failure and diabetic medications. The mean treatment duration was 1.9 ± 1.1 years with carvedilol and 1.4 ± 1.0 years with bisoprolol (p = ns). The carvedilol group achieved a reduction in HbA1c (7.8 ± 0.21% to 7.3 ± 0.17%, p = 0.02) whereas the bisoprolol group showed no change in HbA1c (7.0 ± 0.20% to 6.9 ± 0.23%, p = 0.92). There was no significant difference in the change in HbA1c from baseline to peak BB dose in the carvedilol group compared to the bisoprolol group. There was a similar deterioration in eGFR, but no significant changes in lipid profile or microalbuminuria in both groups (p = ns). Conclusion BB use did not worsen glycaemic control, lipid profile or albuminuria status in subjects with SHF and T2DM. Carvedilol significantly improved glycemic control in subjects with SHF and T2DM and this improvement was non significantly better than that obtained with bisoprolol. BB's should not be withheld from patients with T2DM and SHF.