Temperature is a primary driver to define the ecophysiological activity and performance of ectotherms. Thus, thermal tolerance limits have a profound effect in determining geographic ranges. In regions with extreme cold temperatures, lower thermal limits of species are a key physiological trait for survival. Moreover, thermal niche breadth also plays an important role in allowing organisms to withstand climatic variability and confers species with broader potential to establish in new regions. Here we study the evolution of thermal tolerance limits among Collembola (Arthropoda) and explore how they are affected by the colonization of polar environments. In addition, we test the hypothesis that globally invasive species are more eurythermal than non‐invasive ones. Critical thermal limits (CTmin and CTmax), classic measurements of thermal tolerance, were compiled from the literature and complemented with experimental assays for springtail species. Genetic data of the mitochondrial gene cytochrome oxidase subunit 1 (COI) was used to assemble a phylogeny. Our results show that polar springtails have lower CTmin and lower CTmax compared to species from temperate and tropical regions, consistent with the Polar pressure hypothesis. We found no phylogenetic signal for CTmax, but low values of phylogenetic signal for CTmin. Globally invasive species do not have significantly broader thermal tolerance breadth (CTrange) than non‐invasive ones, thus not supporting the predictions of the Eurythermality hypothesis. We conclude that polar springtails have evolved their thermal niches in order to adapt to extremely cold environments, which has led to decreasing both upper and lower thermal tolerance limits.