This study describes the kinetics of furan and furfural generation in a cake model, for the first time. These process-induced compounds impact safety and sensory aspects of baked products. Understanding their generation with regards to process dynamics will serve food quality design. However, the complexity of real products makes this task challenging. This work provides a novel approach to understand and model chemical reactivity by implementing an inert cake model (starch, water and cellulose), specifically designed for mimicking a sponge cake structure. The addition of reaction precursors (glucose and leucine) to follow Maillard and caramelization reactions, resulted in browning and generated considerable levels of furanic compounds (up to 17.61 ng/g for furan and 38.99 mu g/g for furfural, dry basis). Multiresponse data modeling resulted in a kinetic model which adequately describes experimental concentrations and makes it possible to estimate the degradation of precursors and the behavior of two hypothetic intermediates.