Oppositely charged proteins can form soluble assemblies that under specific physical chemical conditions lead to liquid−liquid phase separation, also called heteroprotein coacervation. Increasing evidence suggests that surface charge anisotropy plays a key role in heteroprotein complexation, and coacervation. Here, we investigated complexation of an acidic protein, β-lactoglobulin (BLG), with two basic proteins, rapeseed napin (NAP) and lysozyme (LYS), of similar net charge and size but differing in surface charge distribution. Using turbidity measurements and isothermal titration calorimetry, we confirmed that LYS binds BLG as expected from previous studies. This interaction leads to two types of phase separation phenomena, depending on pH: liquid−solid phase separation in the case of strong electrostatic attraction and liquid−liquid phase separation for weaker attraction.More interestingly, we showed using dynamic light scattering that NAP interacts with BLG, resulting in formation of assemblies in the nanometer size range. The formation of assemblies was also evident when modeling the interactions using Brownian dynamics for both BLG + NAP and BLG + LYS. Similarly, to DLS, BLG and NAP formed smaller assemblies than BLG with LYS. The molecular details rather than the net charge of BLG and NAP may therefore play a role in their assembly. Furthermore, simulated BLG + NAP assemblies were larger than those experimentally detected by DLS. We discuss the discrepancy between experiments and simulations in relation to the limitations of modelling precisely the molecular characteristics of proteins.