One striking feature of the US health system, for people like us who are interested in evidence on how improvements in the way medical care is provided and financed affect its outcomes and costs, is that we have a pluralistic, not to say fragmented, medical care payment system. What is wrong with fragmentation? Think of a restaurant dinner for a large party of people. Usually they would order salads, main dishes, and desserts from a menu, and might be expected to ask the waiter to calculate the part of the check that represents their dishes – they would pay fee for service – and one could describe the pattern as fragmented. However, what if the group wants to divide the check equally? What if wine is cheaper by the large bottle but diners ordering different entrees want different wines, raising the bar tab? What if it is a restaurant where at least some dishes are better shared than on individual plates? Then a more integrated approach to dining and payment may lower cost may be better – at least for many. Many experts judge an arrangement in which health care is divided individually into different courses and ordered and paid a la carte as a system that is fragmented and ultimately costly to administer and inefficient. That is the challenge for payment reform – to move away from itemized “fee for service” (FFS) pricing to combined payment for a set menu or meal plan, and to do so in a way that will do more good than harm.