By international standards, and especially given its relatively low per capita income,Vietnam has achieved substantial reductions in, and low levels of, infant and under-fivemortality. We review existing evidence and provide some new evidence bearing on thequestion of whether under the economic liberalization program known as Doi Moi thisreduction in child mortality has been sustained. We conclude that it has, but that the gainshave been concentrated heavily among the better-off. As a result, socioeconomicinequalities in child survival are evident in Vietnam—a change from the early 1990swhen none were apparent. We develop survival models to unpack the causes of thisdifferential decline in child mortality, and conclude that a number of factors have been atwork, including reductions among the poor (but not among the better-off) in coverage ofsome health services and in women’s educational attainment. We argue that if theexperience of the late 1990s is to be a guide to the future, the lack of progress among thepoor will jeopardize Vietnam’s chances of achieving the international development goalsfor child mortality. We examine various policy scenarios, including expanding coverageof health services, water and sanitation, and conclude that such measures—whileuseful—will have only a limited impact on the mortality of poor children. We findpotentially large payoffs to programs aimed at narrowing the gap between the poor andbetter-off in the impacts of the various determinants of child survival.