<jats:title>Abstract</jats:title><jats:p>Despite cold adaptation, Antarctic fish show lower growth than expected from the van’t Hoff’s Q<jats:sub>10</jats:sub> rule. Protein synthesis is one of the main energy-consuming processes, which is downregulated under energy deficiency. Considering the effect of temperature on growth performance, we tested if temperature-dependent cellular energy allocation to protein synthesis correlates with temperature-dependent whole-animal growth and thus thermal tolerance. Cell respiration and energy expenditure for protein synthesis were determined in hepatocytes of the circumpolar-distributed Antarctic eelpout <jats:italic>Pachycara brachycephalum</jats:italic> after warm acclimation (0 °C vs 5 °C) and, of two notothenioids the sub-Antarctic <jats:italic>Lepidonotothen squamifrons</jats:italic> and the high-Antarctic icefish <jats:italic>Chionodraco hamatus</jats:italic>. We used intermittent-flow respirometry to analyse cellular response to acute warming from 5 to 10 °C (<jats:italic>P. brachycephalum</jats:italic>) and from 1 to 5 °C (<jats:italic>L. squamifrons, C. hamatus</jats:italic>). Warming-induced rise in respiration was similar between 0- and 5 °C-acclimated <jats:italic>P. brachycephalum</jats:italic> and between <jats:italic>L. squamifrons</jats:italic> and <jats:italic>C. hamatus</jats:italic>. Irrespective of acclimation, warming decreased energy expenditure for protein synthesis in <jats:italic>P. brachycephalum</jats:italic>, which corresponds to reduced whole-animal growth at temperatures &gt; 5 °C. Warming doubled energy expenditure for protein synthesis in <jats:italic>L. squamifrons</jats:italic> but had no effect on <jats:italic>C. hamatus</jats:italic> indicating that <jats:italic>L. squamifrons</jats:italic> might benefit from warmer waters. The species-specific temperature effect on energy expenditure for protein synthesis is discussed to mirror thermal sensitivity of whole-animal growth performance, thereby paralleling the degree of cold adaptation. Clearly more data are necessary including measurements at narrower temperature steps particularly for <jats:italic>C. hamatus</jats:italic> and an increased species’ number per ecotype to reinforce presented link between cellular and whole-animal thermal sensitivity.</jats:p>