Considerable scientific and engineering effort has been directed toward developing medical implant prostheses to replace a large variety of functions within the human body. Despite significant strides to enhance the structural endurance of such devices, e.g., through the use of bioactive coatings to “strengthen” implant/tissue interfaces, few studies have focused on actually predicting the safe life of such prostheses in realistic physiological environments. This is particularly important for many devices, such as cardiovascular stents and safety-critical implants such as heart valves, where mechanical failure in vivo could result in loss of life. In this presentation, the fracture mechanics-based methodology and practice of establishing a rational basis for life prediction and quality control of specific implants is described, with emphasis on the problem of cardiovascular valves and stents. In particular, the implications of choice of material, such as titanium or CoCr vs. pyrolytic carbon in valves, and stainless steels or CoCr vs. Nitinol in stents, are discussed in detail.