Recent evidence demonstrates that large centrosomal coiled-coil (CC) proteins—such as pericentrin, centrosomin, and spd-5—undergo biological liquid-liquid phase separation (PS) in the cell. These large CC proteins contain several CC domains and stretches of disordered regions in between. CC domains are well known for their ability to complex with each other and form a variety of oligomers. Given that CC domains are common in centrosome-related proteins that show PS, and recognizing that the associative nature of CC domains suggests they could contribute multivalent interactions, we are led to ask: Are CC domains themselves sufficient to drive PS of proteins? The link between CC domains and multivalency has not yet been demonstrated largely because the CC proteins that phase separate—which are not explained by other phase-separating mechanisms—are large, disordered, and intractable to work with in vitro. We address this question with molecular simulation using a customizable coarse-grained model of CC proteins. By changing the number of coils, the oligomeric state of the coils, and the interaction specificity between coils, we can obtain a wide variety of behavior ranging from full liquid-liquid PS to distributions of higher order clusters.