Heparin has a high affinity for bone morphogenetic protein-2 (BMP-2), which is a key growth factor in bone regeneration. The aim of this study was to investigate how the rate of release of BMP-2 was affected when adsorbed to nanosized hydroxyapatite (HAP) particles functionalized with heparin by different methods. Heparin was attached to the surface of HAP, either via adsorption or covalent coupling, via a 3-aminopropyltriethoxysilane (APTES) layer. The chemical composition of the particles was evaluated using X-ray photoelectron spectroscopy and elemental microanalysis, revealing that the heparin grafting densities achieved were dependent on the curing temperature used in the fabrication of APTES-modified HAP. Comparable amounts of heparin were attached via both covalent coupling and adsorption to the APTES-modified particles, but characterization of the particle surfaces by zeta potential and Brunauer-Emmett-Teller measurements indicated that the conformation of the heparin on the surface was dependent on the method of attachment, which in turn affected the stability of heparin on the surface. The release of BMP-2 from the particles after 7 days in phosphate-buffered saline found that 31% of the loaded BMP-2 was released from the APTES-modified particles with heparin covalently attached, compared to 16% from the APTES-modified particles with the heparin adsorbed. Moreover, when heparin was adsorbed onto pure HAP, it was found that the BMP-2 released after 7 days was 5% (similar to that from unmodified HAP). This illustrates that by altering the mode of attachment of heparin to HAP the release profile and total release of BMP-2 can be manipulated. Importantly, the BMP-2 released from all the heparin particle types was found by the SMAD 1/5/8 phosphorylation assay to be biologically active.