The kinetics of thermal degradation of ascorbic acid was studied in a Maltodextrin matrix at different temperatures between 75 and 140 degrees C. Isothermal experiments were performed with samples earlier equilibrated in environments of known a(w) at 4 degrees C, enclosed in hermetically sealed vials, with water contents (w) between 0.17 and 1.68 g water/g solids. Sampling times were selected according to ail optimal experimental design in terms of minimum confidence regions of the parameters estimated using the Bigelow model ann estimates from preliminary experiments. The results showed a quadratic dependence of the reference D-value and linear dependence of die z-value with moisture content: D-r(140 degrees C) = 150.07-290.73w+ 269w(2) (min); z = 12.14+22.99w (degrees C) and showed the applicability of D-optimal designs for determining kinetic parameters in complex situations, with limited experimental requirements. The dependence of the rate constants with water content could not be described by the WLF model ill qualitative terms, although in absolute values this model could be risen with constants similar to those expected from glass transition theory The thermodynamic analysis of die results showed a goon application of the compensation theory in the whole range of water contents.