Attention to intracultural diversity in anthropological research has increased, but the implications of that diversity for research design and data analysis in medical anthropology have not proceeded as far. An examination of diversity and its use in guiding data analyses is given here, based on the study of blood pressure and its social and psychological correlates. It is argued that in the specific ethnographic setting of a small West Indian town, social class structures the diversity of the meanings of beliefs and behaviors. Diversity of meanings, in turn, alters the associations of those beliefs and behaviours with blood pressure. Data analyses guided by this orientation demonstrate that the social patterning of blood pressure varies between and within social class. Specifically, it is shown that one model of social and psychological influences on blood pressure applies only to middle-class persons in a small Jamaican community and not to lower-class persons. Medical anthropologists need to be more sensitive to the range of intracultural diversity and to how that diversity can influence the results of research.