Using K326 as materials, flue-cured tobacco with four leaf positions (the fifth, tenth, fifteenth and twentieth leaves) in three sowing dates (May 4^th, May 14^th, and May 24^th) were used in field experiments where parameters of dynamic leaf growth including length and width of leaves were measured. The results showed that: (1) A Richard curve equation,

y=a/(1+e^b-cx)^1/d, was developed and suitable to relationships between leaf growth or growth days and active accumulated temperature; (2) The basic parameter of ultimate growth (a) and growth rate parameter (c) changed slightly, but initial growth parameter (b) and shape parameter (d) changed dramatically. Transplanting dates regulated the equation of the dynamic mainly by controlling parameters b value and d value. The dynamic model could make a good estimation for leaf growth dynamics with the accuracies (0.9966-1.1114), and the precisions (R²) of above 0.9577. The differences reached an extremely significant level; (3) The sequences of growth days with different transplanting date corresponding to leaf length maximum rate showed the following trends: May 4^th > May 14^th > May 24^th. Those of growth amount of the fifth and tenth leaf corresponding to maximum rate of leaf growth showed the same trends, while those of growth amount of the fifteenth and twentieth leaf corresponding to maximum rate of leaf growth showed the opposite trend. Growth days of flue-cured tobacco in rapid growth stage with the same transplanting period are less than those of the initial growth stage and the stable growth stage. Active accumulated temperature of three growth stages among different leaf positions performed the sequence: initial growth phase < rapid growth stage < stable growth stage.