Accurate reference intervals (RI) for commonly measured blood-based analytes are essential for health monitoring programs. Baseline values for a panel of analytes can be used to monitor physiologic and pathophysiologic processes such as organ function, electrolyte balance, and protein catabolism. Our reference population includes 651 serum samples from polar bears (Ursus maritimus) from the southern Beaufort Sea subpopulation sampled in Alaska, USA, between 1983 - 2016. To establish RI for 13 biochemical analytes, we defined specific criteria for characterizing the reference population and relevant subgroups. To account for differences in seasonal life history characteristics, we determined separate RI for the spring and fall seasons, when prey availability and energetic requirements of bears differ. We established RI for five subgroups in spring based on sex, age class, and denning status, and three subgroups in fall based on sex and age class in females. Alkaline phosphatase activities were twice as high in subadult as in adult polar bears in spring (zmales = 4.08, Pmales < 0.001, zfemales = 3.90, Pfemales < 0.001), and did not differ between seasons. Denning females had significantly higher glucose concentrations than non-denning females (z = 4.94, P < 0.001), possibly reflecting differences in energy expenditure during lactation. Ten of the 13 analytes differed significantly between seasons in either males or females; however, the physiologic importance of these differences may be minimal. Establishing these RI allows for temporal monitoring of polar bear health in the southern Beaufort Sea and may prove useful for assessing and monitoring additional polar bear subpopulations in a changing Arctic environment.