Multi-component crystals -salts, co-crystals, or solvates -are usually designed based on the analysis of complementarity of functional groups and intermolecular interactions of the components. However, no crystal design can do without a practical method of crystal growth. Not all compounds that should be 10 expected to exist based on the "synthon approach" can be prepared in real experiments. This paper aims to illustrate that, in addition to the synthon approach, it is equally important to take into account phase diagrams when searching for practical methods of crystallising multi-component crystals, either as single crystals or as fine particles. Here, we describe the crystallization of bis(paracetamol) pyridine solvate from a glass-like metastable phase produced by quench-freezing of the paracetamol-pyridine solution 15 with subsequent low-temperature annealing. These procedures must be carried out strictly within the boundaries of the two-phase region "solid solvate + liquor", which was found only as a result of modelling the phase diagram. The crystal structure was solved by single-crystal X-ray diffraction and compared with co-crystals of paracetamol found in Cambridge Structural Database. The structureforming role of the intermolecular interactions and their characteristics were studied by variable-20 temperature experiments over the range of 100-275 K. This was compared with the structures of pure paracetamol polymorphs and other solvates and co-crystals at ambient and non-ambient conditions.