Aim Heart failure (HF) affects millions of older individuals in both developed and low/middle-income countries. Serum galectin-3 levels have been shown to have prognostic value. However, its use as a diagnostic biomarker has not been explored. The aim was to establish a saliva galectin-3 reference range and to demonstrate the potential diagnostic utility of salivary and serum galectin-3 levels in assessing HF. Methods Blood and saliva samples were collected from age-matched healthy controls (n=51) and patients with HF (n=63). Customised immunoassays were developed to quantify salivary galectin-3 levels. The diagnostic performances of these assays were evaluated by receiver operator characteristic (ROC) curves analysis. Results The galectin-3 concentrations were significantly elevated in saliva and serum samples of patients with HF compared with controls (p<0.001 and p<0.0001, respectively). Using ROC curve analysis, both serum and salivary galectin-3 gave area under the curve (AUC)=0.86 and AUC=0.73, respectively. There was also a significant correlation (r=0.4, p<0.01) between serum and salivary galectin-3 levels. Conclusions For the first time, we have quantified galectin-3 levels in human saliva and have demonstrated potential clinical utility in diagnosing HF. Further, larger multicentre clinical trials are needed before salivary galectin-3 levels can be implemented in a clinical setting.