The life history, genetic, and habitat use differences between the 2 polar bear (Ursus maritimus) subpopulations in Alaska have been used to determine the geographic border separating them, but it has sparked a debate of the correct placement of the border for several years. Recently, the Southern Beaufort Sea (SBS) polar bear subpopulation has declined due to sea ice loss, while the Chukchi Sea (CS) subpopulation appears stable. To provide additional information about potential differences between the SBS and CS subpopulations, such as differences in prey sources, we used stable isotope analysis of carbon and nitrogen from bone collagen of polar bears in these 2 neighboring subpopulations.  We analyzed polar bear bones (1954–2019) from 112 individuals. Our purpose was to determine if the SBS and CS subpopulations could be distinguished based on the stable isotope signatures of bone collagen. A difference greater than 1‰ in δ¹³C values suggests a change in carbon sources, such as nearshore to offshore, while a 3‰ change in δ¹⁵N values equates to a change of about 1 trophic level. Our study showed a significant difference in δ¹³C values (P ≤ 0.001), but not δ¹⁵N values (P = 0.654) between the CS (−13.0±0.3‰ and 22.0±0.9‰, respectively) and SBS bears (−14.7±1.3‰ and 22.2±1.0‰, respectively). Our findings indicate that the 2 subpopulations are consuming similar high trophic level prey, while feeding in ecosystems with different δ¹³C baselines. We performed a logistic regression analysis (LR) using δ¹³C and δ¹⁵N values of the polar bears to predict their placement into these 2 subpopulations. Using Icy Cape, AK as the geographical boundary, the LR correctly placed polar bears in their respective subpopulations 82% of the time. Overall accuracy of placement changed to 84% when using the current geographical boundary at Utqiaġvik, AK. Samples collected from the Wainwright, AK region were predicted as 58% CS and 42% SBS polar bears. This suggests that the area between Wainwright and Icy Cape is a polar bear mixing zone that includes bears from both subpopulations. Bone collagen has a long-term, potentially life-long, stable isotope turnover rate, and our findings could be used to determine the association of harvested polar bears to Alaska subpopulations, thus aiding in transboundary harvest quota management.