The use of exclusionary techniques in the procurement of donors for bone allografts greatly reduces chances for disease transmission. Furthermore, treatment of HIV with either chemical agents or strong acids will effectively inactivate the AIDS virus. These data are taken as indirect proof that the risk of obtaining AIDS from a freeze-dried bone allograft is highly remote. The purpose of this study is to obtain direct evidence that the processing of a demineralized freeze-dried bone allograft would render the allograft safe for human use. In Part I, human cortical bone was obtained from a cadaveric source and tested to be free of HIV contamination. The bone was spiked with 5.26 x 10(9) viral particles. This corresponded to 148 micrograms of total viral protein. In Part II, cortical bone was procured from a donor who died of AIDS. In both Parts I and II, the cortical bone was ground to yield particle sizes of 90 to 500 microns. Test samples were treated with a virucidal agent and demineralized in HCl. Control samples were left untreated. All samples were cocultivated with stimulated peripheral blood lymphocytes and assayed for p24 core protein, reverse transcriptase, and viral gag gene by polymerase chain reaction (PCR). In Part I, the HIV spiking experiment, untreated virus infected particulate bone was positive for HIV replication. Treated samples were negative when assayed for HIV. Bone samples in Part II, HIV infected bone, were positive by PCR. Replication of viable HIV could not be demonstrated after treatment. It was concluded that demineralization and treatment with a virucidal agent inactivates HIV in spiked and infected bone.