An approach to isolate first crystals for molecules prone to liquid-liquid phase separation (LLPS) is described. In the absence of knowledge about the solubility of a hypothetical crystal, it is proposed to determine the LLPS phase boundary, then operate at a temperature of 2 degrees C above the LLPS to maximize the supersaturation with respect to a hypothetical crystalline phase and to avoid LLPS. The proposed approach, combined with a careful solvent and counterion selection based on molecular size, allowed isolating a crystalline salt that was elusive using high-throughput screening. The crystalline salt displayed higher resistance to mechanical stress than the crystalline parent, allowing the progression of the molecule in the development pipeline.