The prediction of implant behaviors in vivo by the use of easy-to-perform in vitro methods is of great interest in biomaterials research. In the field of bone implants, it is commonly accepted that a spontaneous formation of a hydroxyapatite (HA) layer on the implant surface under in vivo conditions is a key step for osseointegration that helps to achieve a direct bone-implant bonding. Based on the hypothesis that the formation of this HA layer can also be reproduced in vitro, immersion test in a Simulated Body Fluid (SBF) was proposed and widely used as an in vitro medium to evaluate the Apatite Forming Ability (AFA) of a material, used as an indicator of its bioactivity in vivo. In view of the limitations of the current SBF, carbonate-buffered SBFs were proposed and the effect of proteins on HA formation was investigated using several titanium-based surfaces in comparison with the current SBF (ISO 23317). The presence of bovine serum albumin (BSA), strongly inhibits the formation of HA in conventional SBF on alkaline-etched titanium with a known in vivo bioactivity, whereas HA is still observed in the case of carbonate-buffered SBFs. The inhibitory effect is also found to be concentration-dependent. The insignificant decrease in solution pH and free calcium concentration due to the addition of BSA suggests other causes, such as surface adsorption, for the inhibitory effect. TiO2 rutile powder was also studied and used to investigate the effect of small amino acids on the nucleation and growth of HA. It was found that at a concentration of 10 mM, the amino acids studied including L-Alanine, L-Serine, L-Arginine, L-Cysteine and L-Lysine had little effect on HA growth. Further studies using non-polar L-Alanine and polar L-Serine revealed a stronger effect from L-serine, which is also concentration-dependent. A new, fast in vitro method based on calcium titration is proposed for the evaluation of in vivo bioactivity. In this method, calcium ions are progressively added to a phosphate solution in contact with the test material until the nucleation of HA occurs. The degree of supersaturation of the solution required for the nucleation is used as an indicator of the bioactivity. Using four titanium-based surfaces, same conclusions were obtained compared to the SBF test. However, this method provides a quantitative result in a matter of a few hours. The bioactive materials tested were observed not only to accelerate nucleation but also to change the crystallization pathway. The conclusions drawn from this innovative method were corroborated by in vitro cell culture tests and in vivo animal experiments on a sheep model conducted by other partners of the project. High-resolution in situ techniques were used trying to investigate the mineralization processes. Cryo-TEM revealed the existence of nanometer-sized species in metastable Tris-buffered SBFs. The crystallization process in highly supersaturated solutions was captured in situ by liquid phase TEM. The single crystal rutile (110) surface was found to show a very low AFA. On the contrary, TiO2 powder with a mixture of rutile and anatase was shown to induce HA formation on its surface. In situ AFM revealed both a direct and an indirect pathway to crystalline calcium phosphate species, highlighting the complexity of the process of heterogeneous nucleation of calcium phosphate on bioactive surfaces.