This paper addresses one of the major obstacles of reaching the Millennium Development Goals (MDG): inefficient and unreliable information systems. Leading international organizations have called for integrated data warehouses as one of the solutions, but this remains hard to achieve. This paper presents four country cases of standardizing and integrating health data which are all following what is here termed a data warehouse approach; data from across different health programs are organized in one database framework – or data warehouse. In all countries, fragmentation of health information in different partly overlapping subsystems run by different vertical health programs represented a major problem for the efficient use of health information. While South Africa developed a new integrated system in addition to the existing fragmented subsystems, Zanzibar, Sierra Leone and Botswana all aimed to encompass all or most of the data from the existing systems. The three latter countries all followed slightly different approaches, more or less incremental in the approach to standardizing health data, and more or less strict in whether to include “all” data, and whether to solve all inconsistencies between the various data sets included early on. The four cases demonstrate that integration is as much, and maybe more, about aligning organizational-political actors as it is about technical solutions. The technical solutions are, however, important in aligning these actors and in enabling integration. We argue that “attractors,” technical solutions or standards that achieve a certain level of success and enable the building of momentum, are important in aligning the various political actors. In turn, these attractors need to evolve within the changing context of a growing health information system in order to achieve the scale needed to address the MDGs with full force.