The basic thesis of this paper is that, in the short to medium term (that is to say, the politically relevant term), fee-for-service payment will be the dominant form of Medicare transaction--and that is not such a bad thing. Capitated arrangements will grow, and should be encouraged, especially for long-term-care services, but for reasons to be discussed, it would be a mistake to put too many policy eggs into that basket. As the nascent Prospective Payment System (PPS) for hospital care reveals, fee-for-service payment can take almost an infinite variety of forms, and the connection between the choice of form and the cost impact is highly variable and somewhat uncertain. PPS needs considerable improvement; so, unquestionably, does the current nonsystem of physician payment. An analysis of potential improvements in both arenas will constitute the bulk of this discussion. That emphasis reflects not only the personal predilections of the author, but also the contention that those are the most sensible things to talk about.