Don Boudreaux, Chairman of the George Mason University Department of Economics, proprietor of Café Hayek, says:

One of my candidates for most underappreciated article ever published in economics is Ronald Coase's "The Marginal Cost Controversy," (Economica, 1946)

Coase's insights are richer than most people suspect, so it's always risky to try to summarize any of this great scholar's articles in a few sentences, but here's my go at this 1946 article:

Economists: loose your devotion to marginal-cost pricing. The best prices are not necessarily those that equal marginal cost. Prices above marginal cost help convey important information - namely, information about the value of the capital invested that makes provision of the good or service possible in the first place. This information, in turn, is important to entrepreneurs searching for profitable places to invest their money and energies.
. . .
Coase's point seems obvious. But it remains shockingly unheeded by economists who, hypnotized from the first with the sterile, static beauty of the (woefully misnamed) "model of perfect competition," are in awe of prices that equal marginal cost.

There's more.

The marginal cost issue is a particular hobby horse of IPcentral, because it is crucial to current debates over intellectual property, ranging from music and movies to games and pharmaceuticals. For prior entries, start here. The Competitive Enterprise Institute has a longer reading list put together for its conference last year on Declining Marginal Cost Industries in the Global Information Age, which was co-sponsored by PFF. The same website has video interviews with economic legends Ronald Coase and Lester Telser.

The Cafe Hayek post has already been picked up by Coyote Blog (Dispatches from a Small Business), which applies it to pharmaceutical prices:

This article helps reinforce the case of why pricing above marginal cost is not necessarily a market "failure". In the case of US Pharmaceutical pricing, drug pricing today is evidence of market failure only if one wishes to see the market fail to develop any new drugs in the future.

And Cyberlibris has posts here and here on marginal cost theory and the intellectual-value economy.