Heat tolerance plasticity is predicted to be an important buffer against global warming. Nonetheless, basal heat tolerance often correlates negatively with tolerance plasticity (‘trade-off hypothesis’), a constraint that could limit plasticity benefits. We tested the trade-off hypothesis at the individual level with respect to heat hardening in two lizard species, Anolis carolinensis and Anolis sagrei. Heat hardening is a rapid increase in heat tolerance after heat shock that is rarely measured in reptiles but is generally considered to be a first line of physiological defense against heat. We also employed a biophysical model of operative habitat temperatures to estimate the performance consequences of hardening under ecologically relevant conditions. Anolis carolinensis hardened by 2 h post-heat shock and maintained hardening for several hours. However, A. sagrei did not harden. Biophysical models showed that hardening in A. carolinensis reduces their overheating risk in the field. Therefore, while not all lizards heat harden, hardening has benefits for species that can. We initially found a negative relationship between basal tolerance and hardening within both species, consistent with the trade-off hypothesis. However, permutation analyses showed that the apparent trade-offs could not be differentiated from statistical artifact. We found the same result when we re-analyzed published data supporting the trade-off hypothesis in another lizard species. Our results show that false positives may be common when testing the trade-off hypothesis. Statistical approaches that account for this are critical to ensure that the hypothesis, which has broad implications for thermal adaptation and responses to warming, is assessed appropriately.