This paper reports on our experiences in setting up and running a health promotion initiative in the south of Manchester, UK. The project was modelled on the work of the travailleuses familiales in France, and consisted of a team of "family workers" who were employed to provide social, practical and psycho logical support to women in pregnancy. The aim of the scheme was to improve the health and well-being of the women and their eventual offspring by reducing stress during pregnancy. The intervention was evaluated in the form of a randomised controlled trial. Those eligible for the trial were women identified as being at above-average risk of giving birth to a low birth weight baby.In this article we present details of the intervention itself rather than of the trial. Aspects of the scheme covered relate to certain issues raised in the course of carrying out the project-for example, the difficulties of introducing a service which uses lay workers and adopts a client-centred approach, ethical dilemmas posed by preventive work, difficulties of cooperation among staff and of coordinating the care offered through various services to women in pregnancy. Following the work of Schwartz, we describe our trial as pragmatic, in an attempt to encompass the additional variables, as described above, which must be taken into account when assessing the impact of any intervention. The research therefore addresses not only the theoretical question of whether social support is effective, but also the pragmatic question of whether social support can be delivered effectively in a "real life" situation.