T. rex teeth indicate it ran as warm as an elephant

Researchers Randon J. Flores and Robert A. Eagle at UCLA applied clumped‑isotope thermometry to enamel from three Hell Creek T. rex teeth—two from a juvenile over three tonnes and one isolated fragment. By measuring the frequency of carbon‑13/oxygen‑18 bonds, they derived temperatures of 37.3 °C, 35.9 °C and 34.7 °C, averaging 36.3 °C ± 2.5 °C. The team first confirmed that the enamel had not been altered over 66 million years, contrasting its isotopic signature with that of the underlying dentin and with modern crocodile enamel. Parallel analysis of contemporaneous crocodilian teeth (≈31 °C) and freshwater mussel carbonates (≈26 °C summer) showed that the theropod was consistently hotter than its environment, a result reinforced by a high‑resolution Late Cretaceous climate model that capped summer peaks at 33 °C.

This work lands squarely in the long‑standing debate over dinosaur metabolism. Earlier oxygen‑isotope studies were confounded by unknown water chemistry, limiting their ability to distinguish endothermy from ectothermy. Clumped‑isotope thermometry sidesteps that problem and has already been used on Jurassic sauropods; extending it to a large theropod provides a rare, direct temperature proxy. The juvenile’s temperature exceeds predictions for a purely inertial “gigantothermy” model, suggesting that metabolic heat production, not just size, kept T. rex warm. The authors’ virtual‑species model, built from thermal tolerances of 465 modern endotherms, projects suitable habitat across almost the entire Late Cretaceous North American landmass, matching known fossil localities and even supporting the plausibility of high‑latitude finds in Alaska and Texas.

If T. rex maintained a stable, near‑elephantine temperature, its energetic needs would have been substantial, reshaping estimates of prey turnover and ecosystem carrying capacity. The study’s limited sample—three teeth from a single formation—means statistical confidence remains modest, and the ±2.5 °C error bar could shift interpretations of metabolic strategy. Future research should broaden the taxonomic sample, apply the same technique to smaller theropods and ornithischians, and integrate the temperature data with biomechanical and ecological modeling to test how a high‑metabolism apex predator fit into Late Cretaceous food webs. Additionally, refining climate simulations and accounting for microhabitat variation will sharpen the habitat‑suitability maps derived from the virtual‑species approach.