Oral processing of two milk chocolates, identical in composition and viscosity, was investigated to understand the textural behaviour. Previous studies had shown differences in mouthcoating and related attributes such as time of clearance from the oral cavity to be most discriminating between the samples. Properties of panellists' saliva, with regard to protein concentration and profile before and after eating the two chocolates, were included in the analysis but did not reveal any correlation with texture perception. The microstructure of the chocolate samples following oral processing, which resembled an emulsion as the chocolate phase inverts in-mouth, was clearly different and the sample that was found to be more mouthcoating appeared less flocculated after 20 chews. The differences in flocculation behaviour were mirrored in the volume based particle size distributions acquired with a laser diffraction particle size analyser. The less mouthcoating and more flocculated sample showed a clear bimodal size distribution with peaks at around 40 and 500 μm, for 10 and 20 chews, compared to a smaller and then diminishing second peak for the other sample following 10 and 20 chews, respectively. The corresponding mean particle diameters after 20 chews were 184 ± 23 and 141 ± 10 μm for the less and more mouthcoating samples, respectively. Also, more of the mouthcoating sample had melted after both 10 and 20 chews (80 ± 8% compared to 72 ± 10% for 20 chews). Finally, the friction behaviour between a soft and hard surface (elastopolymer/steel) and at in-mouth temperature was investigated using a commercial tribology attachment on a rotational rheometer. Complex material behaviour was revealed. Observations included an unusual increase in friction coefficient at very low sliding speeds, initially overlapping for both samples, to a threefold higher value for the more mouthcoating sample. This was followed by a commonly observed decrease in friction coefficient with increasing sliding speed (mixed and elasto-hydrodynamic regime), steeper in the case of the more mouthcoating sample until the differences between the two samples became negligible at a sliding speed of ≈0.2 mm s−1. The stark differences in the tribological behaviour in these regimes begin to allow correlation of data from sensory and physical measurements based on insight into the underlying material behaviour. The complex picture also included comparable behaviour of both samples in the late stages of the elasto-hydrodynamic regime and the early stages of the hydrodynamic regime, until a change of slope was observed and then, at higher sliding speeds, the less mouthcoating sample showed higher friction coefficients. In conclusion, this research uncovered novel correlations of a complex food composite between the sensory behaviour and the physical material properties relating to melting and friction behaviour.