Though the problems translating or applying research in policy-making are legion, solutions are rare. As developing countries increase their capacities to develop effective local solutions to their health problems, they confront the research/policy dilemma. Yet few descriptive studies of research-policy links can be found from developing countries, and the relevance of European and North American models and data is questionable. We report the results of a descriptive study from Mexico of the relationship between health research and policy in four vertical programmes (AIDS, cholera, family planning, immunization). We interviewed 67 researchers and policy-makers from different institutions and levels of responsibility. We analyzed interviewee responses looking for factors that promoted or impeded exchanges between researchers and policy-makers. These were, in turn, divided into emphases on content, actors, process, and context. Many of the promoting factors resembled findings from studies in industrialized countries. Some important differences across the four programmes, which also distinguish them from industrialized country programmes, included extent of reliance on formal communication channels, role of the mass media in building social consensus or creating discord, levels of social consensus, role of foreign donors, and extent of support for biomedical versus social research. We recommend various ways to increase the impact of research on health policy-making in Mexico. Some of the largest challenges include the fact that researchers are but one of many interest groups, and research but one input among many equally legitimate elements to be considered by policy-makers. Another important challenge in Mexico is the relatively small role played by the public in policy-making. Further democratic changes in Mexico may be the most important incentive to increase the use of research in policy-making.