Medical innovations have improved survival and treatment for many diseases but have simultaneously raised spending on health care. Many health economists believe that technological change is the major factor driving the growth of the heath care sector. Whether quality has increased as much as spending is a central question for both positive and normative analysis of this sector. This is a question of the impact of new innovations on quality-adjusted prices in health care. We perform a systematic analysis of the impact of technological change on quality-adjusted prices, with over six thousand comparisons of innovations to incumbent technologies. For each innovation in our dataset, we observe its price and quality, as well as the price and quality of an incumbent technology treating the same disease. Our main finding is that an innovation's quality-adjusted prices is higher than the incumbent's for about two-thirds (68%) of innovations. Despite this finding, we argue that quality-adjusted prices may fall or rise over time depending on how fast prices decline for a given treatment over time. We calibrate that price declines of 4% between the time when a treatment is a new innovation and the time when it has become the incumbent would be sufficient to offset the observed price difference between innovators and incumbents for a majority of indications. Using standard duopoly models of price competition for differentiated products, we analyze and assess empirically the conditions under which quality-adjusted prices will be higher for innovators than incumbents. We conclude by discussing the conditions particular to the health care industry that may result in less rapid declines, or even increases, in quality-adjusted prices over time.