OBJECTIVE To determine for dogs stability of cortisol, thyroxine (T4), and free thyroxine (fT4) in plasma and serum stored in glass or plastic tubes at -20, 4, 25, and 37 C. METHODS Prospective study. METHODS Phase I, 7 Greyhounds; Phase II, 6 mixed-breed dogs. METHODS Phase I: blood was obtained after administration of thyroid-stimulating hormone and adrenocorticotropin. Serum and plasma samples from each dog were divided into 8 aliquots, 4 in glass and 4 in plastic tubes. A pair of aliquots, 1 in plastic and 1 in glass, were stored at -20, 4, 25, or 37 C for 5 days and then assayed for hormones. Phase II: blood was obtained without prior stimulation. For fT4 determination, serum from each dog was placed in plastic or glass tubes, assayed immediately, stored at -20 C for 5 days, and reassayed. Aliquots from each dog were also stored for 1 day at 4 or 25 C and then assayed. Samples for cortisol determination were handled as in phase I. RESULTS Phase I: there was no effect of tube type (glass vs plastic) on cortisol. Cortisol concentrations decreased after storage in serum at 4, 25, and 37 C, and in plasma at 37 C, compared with storage at -20 C. There was no effect of sample type (serum or plasma) on T4. Thyroxine concentrations increased after storage at 37 C in glass, compared with storage at -20 C. The fT4 concentrations were lower in serum than plasma after storage at -20 C. Concentrations of fT4 increased after storage at 37 C in glass, compared with storage at -20 C. Phase II: the fT4 concentrations did not change after storage in any condition. There was no effect of tube type on cortisol concentrations. Serum cortisol concentrations decreased after storage at 37 C, compared with storage at -20 C. CONCLUSIONS For cortisol, cooling of plasma is not necessary, but serum should be shipped cold. For T4 and fT4, serum is sufficient; contained within plastic tubes, samples can be shipped without cooling if assayed within 5 days.