A thin passive titanium dioxide, in its stoichiometric form, has a very high corrosion resistance, but the same conclusion can not be made on corrosion resistance of a surface which is not stoichiometrically titanium dioxide, or even a surface which is a composition of various elements and oxides. In practice, the implants available on the market have an oxide surface contaminated with other elements. The aim of this paper is to correlate clinical observations that show the deterioration of Ti made implants after certain period of insertion in the patients, and in vitro corrosion resistance of Ti implants with surface passive oxide layer. For this purpose, surface analysis of the retrieved failed implants were performed and in vivo animal experiments with relation to ion release from implants were done. Finally, on the basis of the clinical observation, in vivo animal test, and in vitro electrochemical corrosion test, a model is proposed to explain the corrosion and ion release from the Ti implant.