Both citrate ions and proteins with phosphorylated side chains are believed to play central roles in bone mineralization. Using metadynamics simulations, we elucidated the pH-dependent surface binding of citrate and O-phospho-L-serine (Pser) at structurally disordered (100) and (001) surfaces of Ca hydroxyapatite (HA), the mother structure of bone mineral. The binding strength of Pser at the (100) surface increased concurrently with the pH value from 4.5 to 14.0 and remained consistently stronger than that at the (001) surface. In contrast, citrate revealed very similar adsorption affinities at both (100) and (001) surfaces throughout 7.4 <= pH <= 12.3, whereas adsorption at the (100) surface was favored at the lowest (4.5) and highest (14.0) pH values. The two most stable/probable binding modes of citrate involved either a simultaneous anchoring of all three COO- groups such that the molecule caps the HA surface or a tilted configuration stemming from the dual binding of the central and one terminal COO- moiety. The surface adsorption of Pser is dominated by its phosphate group, which participates in all significant binding modes, with the two most prominent ones featuring either a co-binding of the PO4 and COO- sites or a linear alignment of the molecule with the surface by the simultaneous anchoring of all three phosphate, carboxy, and amino groups. The latter constitutes the most stable binding mode at the (100) surface for pH = 7.4. We also introduced a straightforward analysis protocol based on Debye-Hiickel energies, enabling the quantification of the relative contributions of each functional group of an adsorbed molecule, as well as its underlying interaction energies with the surface. Both the Pser and citrate adsorption occurred predominantly through electrostatic Ca2+-COO-/PO42- interactions, along with overall minor H-bond contributions, except for the Pser binding at the PO4-richer (001) HA surface for pH values between 7.4 and 12.3. We highlight the importance of excluding OH groups in the HA surface model to better mimic real nanocrystalline apatite particles and improve the accuracy of the adsorption modeling.