Leaf thermoregulation and consequent leaf-to-air temperature difference (ΔT) are tightly linked to plant metabolic rates and health. Current knowledge mainly focus on the regulation of environmental conditions on ΔT, while an accurate assessment of biotic regulations with field data remains lacking. Here, we used a trait-based model that integrates a coupled photosynthesis-stomatal conductance model with a leaf energy balance model to explore how six leaf traits (i.e. leaf width, emissivity, visible and near-infrared light absorptance, photosynthetic capacity—Vc,max25, and stomatal slope—g1) regulate ΔT variability across the diel timescale. We evaluated the model with field observations collected from temperate to tropical forests. Our results show that: (1) leaf traits mediate large ΔT variability, with the noon-time trait-mediated ΔT variability reaching c. 15.0 °C; (2) leaf width, Vc,max25, and g1 are the three most important traits and their relative importance in ΔT regulation varies strongly across the diel timescale; and (3) model-derived trait-ΔT relationships match field observations that were collected close to either midday or midnight. These findings advance our understanding of biotic controls of leaf-level ΔT variability, highlighting a trait-based representation of leaf energy balance that can improve simulations of diverse leaf thermoregulation strategies across species and physiological responses to climate change.