In this work, the specific effect of microwave (MW) energy on chemical reactions was investigated by online monitoring of the decomposition kinetics of potassium persulfate (K2S2O8). Experiments conducted at constant temperature and constant MW power revealed that the rate constant was about 1.1-1.8 times higher than the rate constant of the thermally heated system at the same temperature, depending on the MW power. To model the dependence of the rate constant on the MW power as k = f(P) exp(-g(P)/T), various functional forms of MW power were envisaged for g(P) (activation energy E-a) and for f(P) (preexponential factor k(o)). Linear, quadratic, and cubic polynomial models were tried for both functions. Nonlinear regression analysis was performed and detailed statistical analysis was applied, and the best results were obtained with quadratic f(P) and cubic g(P) functions, with the highest R-adj(2) (0.9975) and the lowest standard deviation (0.94 x 10(-5)). The mathematical model revealed that both the activation energy and preexponential factor were higher than their thermal counterparts (for P = 0.75 kW dm(-3), k(o,mw)/k(o,th) = 120, E-a,E-mw/E-a,E-th = 1.12).