A complete kinetic modeling of the polymer-derived ceramics (PDCs) route is achieved for the first time through the investigation of the solid-state decomposition of a typical melt-spinnable poly[B-(methylamino)borazine] into boron nitride fibers at various heating rates. Through the use of the Lorentz fitting approach, it is shown that the two-step weight loss associated with the polymer-to-ceramic conversion is governed by a complex interplay of five diffusion-type transport mechanisms that are independent of the applied heating schedule. The application of the Friedman method to dynamic thermogravimetry data yields E-a and ln A values that are seen to increase with the extent of the ceramic conversion from region one (E-a = 38.73 kJ mol(-1)) to region five (E-a = 146.64 kJ mol(-1)). This fact indicates that both the mechanisms within those regions are parallel routes to the formation of the final solid-state material and their complexity increases with the conversion progress. The cross-linking process (first weight loss) of the polymer is governed by three dependent poorly energetic mechanisms. The first weight loss is activated by ammonolysis reactions that provide a modified polymer capable of undergoing condensation reactions in regions two and three to yield a highly cross-linked polymer. A large evolution of methylamine is identified during this process. Mineralization (region four) and ceramization (region five) steps are represented by two highly energetic multistep mechanisms. The mineralization step is associated with a large evolution of methylamine and occurs during the transition between the cross-linking and ceramization processes through the cleavage of the interring B-N bonds. Ceramization represents the end of the polymer-to-ceramic conversion in which the planar consolidation of BN hexagons occurs through complex structural rearrangements of the borazine units (cleavage of the intraring B-N bonds) accompanied with an ammonia evolution. Finally, the simulation of the polymer-to-ceramic conversion was demonstrated through a simplified model that appropriately predicted experimental data.