Abstract: Rapidly and accurately predicting leaf area (LA) and leaf dry mass (LDM) are essential for exploring the response of plant traits to climate change. Empirical models suitable for predicting LA and LDM of a single leaf for various broadleaved tree species at the regional scale have not been proposed. We selected six broadleaved tree species in four mixed broadleaved-Korean pine (Pinus koraiensis) forests in northeastern China, including Betula platyphylla, Tilia amurensis, Populus davidiana, Betula costata, Fraxinus mandshurica and Ulmus laciniata, and measured leaf length, leaf width, leaf thickness, LA and LDM at different canopy layers (top, middle, and low). Using the median of leaf length and width ratio as the classification criterion, the six species were sorted into two groups. We tested whether different canopy layers for each group of broadleaved tree species had significant impacts on the empirical model for predicting LA and LDM. We constructed empirical models suitable for predicting LA and LDM of a single leaf at different canopy layers at the regional scale, and verified their forecast accuracy, and further evaluated their applicability for predicting LA and LDM of same broadleaved tree species in other regions. These results showed that the LA of a single leaf increased significantly with the decreases of canopy height for the six tree species, while the LDM of some broadleaved tree species showed a downward trend. The canopy height had significant impacts on constructing the empirical model for LA and LDM. The average forecast accuracy of empirical model was 95% and 83% for LA and LDM of a single leaf across canopy layers for two groups of broadleaved tree species, respectively. The average forecast accuracy was 94% and 80% for predicting LA and LDM of corresponding broadleaved tree species in other regions, respectively, indicating that the empirical models constructed in this study had a universal applicability in Northeast China.