A theoretical model was developed that explains why the erosion of biodegradable bulk-eroding polymers follows the same kinetics on different time scales. poly(D,L-lactic acid-co-glycolic acid) 1:1 of various molecular weight served as a model compound. After a period of no mass loss, those polymers erode by spontaneously losing more than half of their mass, which is typical for bulk-eroding polymers in general. Until now this was explained with degradation accelerated by auto catalysis inside the bulk compared to the surface of polymer matrices. To investigate if other mechanisms might be involved as well, an erosion model was developed. It assumes that erosion can start only after the polymers are degraded to water-soluble products and after these degraded polymer areas have contact to the erosion medium. Simulations revealed that percolation phenomena may be responsible for the spontaneous mass loss and that the time of onset is a function of the degradation velocity. Degradation rate constants determined from fits to experimental data of the investigated poly(alpha-hydroxy esters) agree well with literature values for ester hydrolysis. This suggests that the model is in agreement with existing degradation theories.