Past studies suggest that rates of woody plant resprouting following a "topkilling" disturbance relate to timing of disturbance because of temporal patterns of below-ground carbohydrate storage. Accordingly, we hypothesized that fire-return interval (1 or 2 years) and season of burn (late dormant or early growing season) would influence the change in resprout growth rate from one fire-free interval to the next (Δ growth rate) for broadleaf woody plants in a pine-grassland in Georgia, USA. Resprout growth rate during one fire-free interval strongly predicted growth rate during the following fire-free interval, presumably reflecting root biomass. Length of fire-free interval did not have a significant effect on mean Δ growth rate. Plants burned in the late dormant season (February-March) had a greater positive Δ growth rate than those burned in the early growing season (April-June), consistent with the presumption that root carbohydrates are depleted and thus limiting during spring growth. Plants with resprout growth rates above a certain level had zero or negative Δ growth rates, indicating an equilibrium of maximum resprout size under a given fire-return interval. This equilibrium, as well as relatively reduced resprout growth rate following growing season fires, provide insight into how historic lightning-initiated fires in the early growing season limited woody plant dominance and maintained the herb-dominated structure of pine-grassland communities. Results also indicate tradeoffs between applying prescribed fire at 1- versus 2-year intervals and in the dormant versus growing seasons with the goal of limiting woody vegetation.