Rising temperatures pose a significant threat to the survival of ectothermic animals. Thermal physiological traits, including upper thermal limits, are proven useful traits to assess the vulnerability of ectotherms to changing temperatures. For instance, one may use upper thermal limits to estimate current and future thermal safety margins (i.e., the proximity of upper thermal limits to experienced temperatures), use this trait in coercion with other physiological traits in species distribution models, or investigate the plasticity, heritability and evolvability of these traits for buffering the impacts of changing temperatures. While datasets on thermal tolerance limits have been previously compiled, they often report single estimates for a given species, do not present measures of data dispersion, and are biased towards certain parts of the globe. In this project, we systematically searched the literature in seven languages and produced a dataset of over 3000 heat tolerance estimates for 616 amphibian species. I will first present how the dataset we compiled can be used to address various questions in ecophysiology, and its limitations. Indeed, despite its scope, this dataset also suffered from geographical biases and captured less than 10% of described amphibian species. Here, I will demonstrate how we can fill the missing parts of this puzzle using powerful imputation procedures and estimate the heat tolerance of most described amphibian species. I will then demonstrate how we used this comprehensive database to estimate the vulnerability of the worlds’ amphibians to current and projected temperatures.