The objective of this study was to investigate the mechanisms underlying drug release from a controlled colonic release (CCR) tablet formulation based on a xyloglucan polysaccharide matrix and identify the factors that control the rate of release for the purpose of fundamentally substantiating the concept and demonstrating its robustness for colonic drug delivery. Previous work demonstrated in vitro limited release of 5-aminosalicylic acid (5-ASA) and caffeine from these tablets in small intestinal environment and significant acceleration of release by xyloglucanase, an enzyme of the colonic microbiome. Targeted colonic drug delivery was verified in an animal study in vivo. In the present work, interaction of the xyloglucan matrix tablets with aqueous dissolution media containing xyloglucanase was found to lead to the spontaneous formation of a hydrated highly viscous gummy layer at the surface of the matrix which had a reduced drug content compared to the underlying regions and persisted with a nearly constant thickness that was inversely correlated to the enzyme concentration throughout the duration of the release process. Enzymatic hydrolysis of xyloglucan was determined to take place at the surface of the matrix leading to matrix erosion and a relation for the rate of enzymatic reaction as a function of bulk enzyme concentration and the concentration of dissolved xyloglucan in the gummy layer was derived. A mathematical model was developed encompassing aqueous medium ingress, matrix metamorphosis due to xyloglucan dissolution and matrix swelling, enzymatic hydrolysis of the polysaccharide and concomitant drug release due to matrix erosion and simultaneous drug diffusion. The model was fitted to data of reducing sugar equivalents in the medium reflecting matrix erosion and released drug amount. Enzymatic reaction parameters and reasonable values of medium ingress velocity, xyloglucan dissolution rate constant and drug diffusion coefficient were deduced that provided an adequate approximation of the data. Erosion was shown to be the overwhelmingly dominant drug release mechanism while the role of diffusion marginally increased at low enzyme concentration and high drug solubility. Changing enzyme concentration had a rather weak effect on matrix erosion and drug release rate as demonstrated by model simulations supported by experimental data, while xyloglucan dissolution was slow and had a stronger effect on the rate of the process. Therefore, reproducible colonic drug delivery not critically influenced by inter- and intra-individual variation of microbial enzyme activity may be projected.