A kinetic model for solid-state polycondensation of nylon-6 was developed by extrapolation of the melt chemistry to the amorphous phase of the solid polymer. In this way, a phenomenological description of the process was obtained that required examining al1 possible influences of the semicrystalline structure on the kinetics, equilibrium, and transport phenomena in the polymer granules. Particularly, dramatic apparent kinetics and equilibria changes were postulated due to the confinement of reactive species in a smaller reaction volume upon crystallization. In addition, the dimensional constraint applied by the micro-morphology on the intrinsic kinetics was also assessed. The effect of crystallinity on molecular transport was examined for both the migration of by-products at millimeter scale and the diffusion of reactive functionalities at nanometer scale. In the latter case, the occurrence of a gel effect in step growth polymers was questioned based on the implication of interchange reactions. A detailed reaction scheme was formulated that incorporated the presence of terminated polymer chains, which can result from the use of chain regulators or from degradation during the melt prepolyrnerization stage. Predictions were made for the characteristic quantities of the output product, such as molar mass, polydispersity index, concentration of end-groups, water content, extent of remonomerization, and formation of the cyclic dimer. A single particle model was validated by means of batch experiments carried out in a battery of fixed-bed reactors at laboratory scale. Both the kinetics of polymerization and the evolution of the morphology were tracked, based on techniques including viscometry, titration of functionalities, chromatography, X-ray diffraction, calorimetry, and polarized light microscopy. This led to a refined picture of the complex interplay between reaction and structure. Particularly, implications of the rigid amorphous phase, meso-scale spherulitic structures, polymer skinning, or of the presence of a crystallinity gradient within the polymer were underscored. In addition, the performance of the solid-state model was extensively tested with regard to literature data. A scale-up of the single particle model was performed to deal with the dynamic operation of an industrial moving-packed bed reactor. Simulations allowed the prediction of reactor start-up and shutdown, possible fluctuation in the feed composition, and on-line grade transition. Potential implications of the developed model as a tool for on-line reactor optimization and process control were discussed, based on a parametric sensitivity analysis.