A time-resolved immunofluorometric assay (TR-IFMA) was developed for the determination of haptoglobin (Hp) in canine serum. Haptoglobin was purified from canine acute phase serum by ammonium sulphate precipitation followed by gel filtration. This isolated dog Hp was used as the standard to calibrate the assay. Intra- and inter-assay coefficients of variation of the assay were, respectively, 5.7% and 16.6% at 0.51 mg/mL, 2.4% and 10.6% at 2.1 mg/mL and 10.5% and 11.9% at 32.5 mg/mL. The dilution of serum samples with high Hp concentrations resulted in linear regression equations with R2 of 0.99 and 0.97. A high correlation was found in serum Hp measurements by TR-IFMA and a commercial assay based on peroxidase activity of haemoglobin bound to haptoglobin (R2 = 0.96). The limit of detection for the TR-IFMA method was 0.002 microg/mL. The addition of fresh haemolysate to serum samples did not affect the haptoglobin concentration (P = 0.694). Statistical differences (P < 0.003) were found between healthy dogs and dogs with different pathological processes. In whole blood, Hp concentrations were much lower than in serum but closely related (R2 = 0.84) whereas saliva Hp concentrations were poorly related with serum concentrations (R2 = 0.53). However, the concentration of Hp in saliva was significantly (P < 0.039) higher in dogs with pathological processes compared to healthy dogs. The assay sensitivity was adequate to also be applied to whole blood and saliva specimens.