The distinction between wanted and unwanted fertility has been crucial in many of the more intense debates in recent decades over the nature of contemporary fertility declines and, in particular, the potential impact of expanded provision of family planning services. In a much-debated article published in 1994, Pritchett argues that decline in desired fertility is overwhelmingly the principal source of fertility decline, with the implication that family planning programmes are of little consequence. I revisit this debate drawing on a far larger body of survey data and, more importantly, an alternative fertility specification which relies on a non-conventional definition of wanted and unwanted fertility rates and which distinguishes rates and composition. Decompositions of fertility decline in the period from the mid-1970s to the present are carried out for 44 countries. The decomposition results indicate that declines in unwanted fertility rates have been at least as important, if not more important, than declines in wanted fertility rates. Surprisingly, shifts in the proportion of women wanting to stop childbearing—i.e., changes in preference composition—has contributed very little to fertility change in this period. Further, decline in wanted fertility and increases in non-marital exposure (due largely to delayed entry into first marriage) have also made substantial contributions, although on average they fall short of the contribution of declines in unwanted fertility rates. That declines in unwanted fertility have been an essential feature of contemporary fertility decline is the main conclusion from this research. This in turn opens the door to new perspectives on fertility pre-, mid-, and post-transition which recognises the inter-dependencies between fertility demand and unwanted fertility rates in the determination of the overall level of fertility.