The dynamic relationship between unfolding leaves and active accumulated temperature (AAT) of different maize (Zea mays L.) varieties was studied under different sowing dates, which would provide a theoretical basis for the realization of the informatization and digitalization of maize production on the Loess Plateau of China Results showed that as the number of unfolded leaves increased, the number of days and the required active accumulated temperature to unfold one leaf showed a single peak trend. The unfolding of the 1st and the last leaves need the least time and AAT, while the 8th -15th leaves need the most. With the delay of sowing date, the peak value of the variation trend of the days and active accumulated temperature to unfold one leaf basically decreased gradually. Among different fitting models, the 3rd degree polynomial fitting model y = a + bx + cx(2) + dx(3) may have better biological significance and was thus adopted. As a result, the established models showed great variations in c and d parameters for different varieties, a and c for different sowing dates. The model was tested and showed that the simulated values were in good agreement with the actual values. The above results can provide valuable reference for the quantitative leaf unfolding and AAT requirement for maize and its adaptation to climate change.