Glucose determination is used for diagnosis and prognosis of several diseases (diabetes mellitus, liver diseases, babesiosis) and for screening and monitoring in many different studies (nutritional, therapeutic). The increase of disease cases, especially canine diabetes mellitus, in veterinary clinical practice demonstrates the need to provide new methods more reliable than measure exclusively glucose for its diagnosis. Glucose level can drastically vary due to stress or ingestion of food before blood collection and consequently affect the correct interpretation of the real cause of hyperglycemia. Besides the increase in blood glucose, it's also observed an elevation of fructosamine and glycosylated hemoglobin in patients with diabetes mellitus. Fructosamine is a serum protein formed by an irreversible and non-enzymatic reaction between two molecules of glucose and protein. The albumin, the main protein involved, has a half-life of approximately 8 days in dogs, therefore fructosamine reflects the blood glucose for that period. The glycosylated hemoglobin (GHb) has its origin in the glycosylation of hemoglobin by a slow, irreversible and non-enzymatic reaction. Its concentration depends on the half-life of erythrocytes (120 days) and the concentration of glucose in plasma, meaning that the concentration of GHb remains stable for a period of 2-3 months. It is important that each laboratory has its own reference values and it's known that physiological changes occur in the concentration of analytes in different populations according to factors inherent to the animal (breed, age and gender) and external factors (climate, geography and dietary habits). The objective of this study was to establish reference values for fructosamine and glycosylated hemoglobin in clinically healthy dogs from Porto Alegre. Blood samples were collected into three different tubes (EDTA-K2, NaF/EDTA-Na2 and no-additive) from 87 dogs of different large breeds, including German Sheppard (n = 21), Rottweiler (n = 14), Labrador (n = 4), Dogo Argentino (n = 17), Great Dane (n = 15) and Golden Retriever (n = 16). EDTA whole blood was used for GHb measurement, NaF/EDTA-Na2 for glucose and serum samples for fructosamine. The obtained results were 104.68 ± 24.10 mg/dL for glucose, 252.13 ± 2.84 mmoL/L for fructosamine and 8.03 ± 2.22% for GHb. Statistical difference was found between breeds only for glucose and no statistical difference was found between genders (p < 0.05). The concentrations of all metabolites studied were within the range reported in the literature. The authors conclude that these results may help the clinicians during clinical routine but more studies are needed, especially in small canine breeds.