<p>A conventional explanation for the existence of species-specific maximum growth temperature (MGT) is the occurrence of irreversible loss of essential cellular functions. Hence the hypothesis that the MGT is the thermal point at which cellular destructive processes outpace constructive processes. For a data-based assessment of this hypothesis, we develop a trait-based model of cellular growth in which key cellular traits are the activation energies of constructive and destructive processes. Using Bayesian inversion on growth curves of archaea, we infer trait values and map them to maximal growth temperatures. We identify the difference between those activation energies as a primary driver of variation in maximum growth temperature. The known yet unexplained correlation between maximal and optimal growth temperatures appears to be underpinned by a linear scaling of these activation energies. This scaling relationship points to the plausibility of adaptation to temperatures exceeding the currently known upper limit (110-120 °C) for microbial growth.</p>