Although consumer demand for plant-based cheese alternatives is increasing over the years, their market share remains limited due to poor meltability and texture performance compared to dairy cheeses. In this study, we investigated how compositional and structural differences between dairy and plant-based cheeses govern their thermal and rheological behaviour, with the goal of informing future product innovation. By combining temperature-dependent rheology and temperature-modulated differential scanning calorimetry (TMDSC), we demonstrate that these complementary techniques can effectively differentiate true melting from softening, revealing fundamental differences in structure–function relationships across cheese matrices.

We systematically compared commercially available dairy cheeses and plant-based cheese alternatives across a range of formats, including different ripening stages, processed cheeses, and plant-based products in block and grated forms. Viscoelastic and structural responses were assessed using small- and large-amplitude oscillatory shear (SAOS and LAOS) measurements, supported by confocal and polarised light microscopy. Thermal behaviour was evaluated using TMDSC to determine the freezable water content and the thermal transitions upon heating. This analysis separated reversing and non-reversing heat flows, a new method in the context of this research. Additionally, meltability (modified Schreiber test), water-holding capacity, and oil release were measured. This multimodal methodology enabled direct linkage of composition and matrix organisation to texture and melting behaviour, providing mechanistic insight into performance differences between dairy and plant-based cheese matrices.

Plant-based samples were found to be primarily starch–tropical plant fat matrices, with minor amounts of protein, while dairy samples consisted of protein-stabilised emulsion-filled gels. Temperature-dependent rheology separated the products into four distinct groups, classified based on the reversibility of changes in the moduli and the occurrence of true melting as opposed to simple softening. To our knowledge, this is the first systematic classification of meltability behaviour in these matrices. The four groups were defined as: (1) softening without reformation, (2) softening with reformation, (3) melting, and (4) collapse. True melting below 95 °C was observed predominantly in dairy cheeses and in only one plant-based block product, characterised by a sharp decrease in storage modulus (G′) and a clear G′–G″ crossover. In contrast, most plant-based cheeses exhibited a gradual reduction in G′ without crossover, consistent with progressive softening rather than network disintegration. The corresponding thermal transitions, as observed by TMDSC, were dominated by the non-reversing component of the heat flow, suggesting that the thermal behaviour of these cheeses was primarily governed by irreversible structural rearrangements rather than reversible melting processes. Together, the complementary rheological and TMDSC analyses demonstrate that visual melting does not necessarily correspond to true rheological melting, providing mechanistic insights into the distinct structural dynamics of dairy and plant-based cheese matrices.

Furthermore, the confocal scanning laser microscope (CLSM) showed all samples to be emulsion-filled hydrogels, protein networks in dairy cheeses versus starch matrices in plant-based alternatives. Variations in fat globule size and distribution, however, did not explain fat release. By systematically linking composition, microstructure, rheology and thermal transitions, we provide a detailed account of the techno-functional properties that still need to be improved to better match consumer experience.