Polymer-induced liquid precursor (PILP) phases of calcium carbonate have attracted significant interest due to possible applications in materials synthesis, and their resemblance to intermediates seen in biogenic mineralisation processes. Further, these PILP phases have been formed in vitro using polyelectrolytes such as poly(aspartic acid) which bears many structural parallels to the highly acidic biomacromolecules that are associated with biogenic calcium carbonate. This article describes experiments which investigate how the composition of acidic polypeptides determines their ability to form PILP phases of CaCO3, and therefore whether it is feasible that the acidic biomacromolecules extracted from CaCO3 biominerals could also function in this way. A series of random copoly(amino acid)s constructed from 80–20%, 50–50% and 20–80% aspartic acid and serine residues were synthesised and their effect on CaCO3 precipitation was determined. A strong correlation between the composition and function of the polypeptide was observed. Only the polypeptide containing 80% aspartic acid residues (Asp80%–Ser20%) induced the formation of continuous CaCO3 films, which provide a fingerprint of an intermediary PILP phase, while addition of Mg2+ also facilitated the formation of expanded film-like structures with the polypeptide Asp50%–Ser50%. In contrast, the weakly-acidic polypeptide Asp20%–Ser80% had only a minor effect on the crystal morphologies and also failed to aid infiltration of CaCO3 into small pores. These results therefore demonstrate that counter-ion induced phase separation of highly acidic biomacromolecules proteins appears to be entirely feasible based upon their composition, but that evidence for the operation of this mineralisation mechanism in vivo is still required.