The World Health Organization's challenge of `Health for all by the year 2000' has been associated with a number of campaigns aimed at reducing child mortality. In this area there is both good news and bad news. On the one hand there have been extremely impressive reductions in child mortality in many countries. On the other, some of the declines have stalled in recent years, levels of child mortality remain unacceptably high in many parts of the world, and declines have masked an often widening gap between rich and poor. One of the important initiatives not discussed in this theme issue is the Expanded Programme on Immunization (EPI). This WHO programme was launched in 1974. By 1995, 80% of the world's children had been immunized against six major infectious diseases compared to only 5% in 1974. Top-level political support, social mobilization, and immunization days all helped to increase demand and to facilitate efficient service delivery. The demographic impact, though hard to measure, is likely to be high. Our recent simulations (1) suggest that in countries where mortality, has been high, EPI could over the half century from 1975 have increased life expectancy by over four years and led, for example, to a 33% decrease in female child mortality. This is great news, and might tempt one to think that the global community, should now move on to its next major problem, leaving immunization as a sustainable health service. That would be a catastrophic mistake. The impressive levels of coverage are not uniform, and the immunized do not represent a random sample from their respective populations. In Africa and Central Asia high proportions of babies remain unimmunized and in other areas coverage is limited mainly to the rich. In other words the health services are failing those most in need. This may call for a reorientation from mass campaigns towards targeted ones, so as to maintain the overall decline in child mortality while reducing the unacceptable gap between rich and poor. Improved delivery of health services requires decent data on which to base policy. Surveillance systems have been built up in recent years to provide such data, and these have been supported by excellent large-scale survey programmes such as the Demographic and Health Surveys. However, there are large numbers of people in the world who are not covered by these impressive data collection instruments. However good a census is, it often has difficulty in covering the poorest communities, as these are are often the least accessible. In addition, poor communities grow quickly, and are therefore likely not to feature adequately in the sampling frames from which the large-scale surveys are drawn. Thus people for whom the delivery, of affordable health services is both lacking and massively needed are simply not covered by the available data, and are missing from the global map. Stringent and targeted efforts are needed to obtain information on their position and needs, so that appropriate interventions can be focused on them. …