In this paper we raise questions concerning the widely accepted proposition that economic efficiency requires that copyright protection be limited in its duration (often shorter than the current term). We show that just as an absence of property rights in tangible property would lead to inefficiencies, so intangible works that fall into the public domain may be inefficiently used because of congestion externalities and impaired incentives to invest in maintaining and exploiting these works. Although a system of indefinite renewals could lead to perpetual copyrights or very long terms, this is unlikely. Our empirical analysis indicates that (1) fewer than 11 percent of the copyrights registered between 1883 and 1964 were renewed at the end of their 28-year term, even though the cost of renewal was small; (2) copyrights are subject to significant depreciation and have an expected or average life of only about 15 years; and (3) copyright registration and renewals are highly responsive to economic incentives for the shorter the expected life of a copyright and the higher the registration and renewal fees, the less likely are both registration and renewal. This in turn suggests that a system of modestly higher registration and renewal fees than at present, a relatively short initial term (20 years or so), and a right of indefinite renewal (possibly subject to an overall maximum term of protection of say 100 years) would cause a large number of copyrighted works to be returned to the public domain quite soon after they were created. A further benefit of indefinite renewal is that it would largely eliminate the rent-seeking problem that is created by the fact that owners (and users) of valuable copyrights that are soon to expire will expend real resources on trying to persuade (dissuade) Congress to extend the term.