Public health and its ''basic science'', epidemiology, have become col
onised by the individualistic ethic of medicine and economics. Despite
a history in public health dating back to John Snow that underlined t
he importance of social systems for health, an imbalance has developed
in the attention given to generating ''social capital'' compared to s
uch things as modification of individual's risk factors. In an illustr
ative analysis comparing the potential of six progressively less indiv
idualised and Inure community-focused interventions to prevent deaths
from heart disease, social support and measures to increase social coh
esion fared well against more individual medical care approaches. In t
he face of such evidence public health professionals and epidemiologis
ts have an ethical and strategic decision concerning the relative effo
rt they give to increasing social cohesion in communities vs expanding
access for individuals to traditional public health programs. Practit
ioners' relative efforts will be influenced by the kind of research th
at is being produced by epidemiologists and by the political climate o
f acceptability for voluntary individual ''treatment'' approaches vs u
niversal policies to build ''social capital''. For epidemiologists to
further our emerging understanding of the link between social capital
and health they must confront issues in measurement, study design and
analysis. For public health advocates to sensitise the political envir
onment to the potential dividend from building social capital, they mu
st confront the values that focus on individual-level causal models ra
ther than models of social structure (dis)integration. The evolution o
f explanations for inequalities in health is used to illustrate the na
ture of the change in values. (C) 1998 Elsevier Science Ltd. All right
s reserved.