In recognition of the considerable importance given in the literature to social capital as an instrument for both growth maximization and poverty reduction, this survey was launched as a pilot attempt to assess the contribution of microfinance to community-building and political participation, with a view to designing an instrument which might be useful for the monitoring systems of microfinance organizations. Empirical work is carried out in Russia, Slovakia and Romania through two NGOs: the INTEGRA Foundation and the FORA Fund. We acknowledge at the beginning considerable problems of methodology: in particular of causation between microfinance and social capital variables, and of definition of the social capital variable itself. The contrast between ‘associational’ and ‘trust’ definitions of social capital turns out to be particularly relevant in the present case. Results: impact. In all countries examined, there appears from the questionnaires to be relatively little impact of microfinance on associational membership; the main difference between treatment and control samples appears to be a higher membership of church associations, both in Slovakia and Russia, amongst borrowers. Interview data, however, suggest a much stronger link than appears from the questionnaire, in particular an influence from the provision of finance to informal political organization, beginning with extensions of the solidarity group model, but also found among individual clients. These informal organizations in turn are of three types, (i) bonding within groups for social or ‘political’ purposes, (ii) bonding between groups (called by Woolcock ‘bridging social capital’), (iii) bonding between individuals or groups and higher-level hierarchical organizations such as branches of local government or commercial banks. (i) was very important within solidarity and trust groups, and was not classified by the questionnaire as associational membership. There was a dramatic illustration of (ii) in Borisoglebsk, Russia, where some solidarity groups had come together in a struggle against local authority corruption; (iii) was found in Slovakia and also in Novgorod, Russia, and in general was found at higher levels of income and confidence. There are also (in Slovakia, the only country so far examined) considerable differences between microfinance clients and others in terms of perceptions of corruption and its influence on their business. We are currently investigating the determinants of informal association by regression methods, and speculate that as well as experience, leadership, intragroup inequality, income and education may be very important. Possibly stemming from these forms of association, there was also a significantly higher level of trust (on average and in particular agencies) amongst microfinance customers than the control group, particularly towards agencies of local government.1 Thus to summarise microfinance scarcely influences formal associational membership, but greatly influences informal behaviour patterns, thence trust, and thence informal political participation. There are three dimensions of this relationship which are particularly interesting. The first is the influence of prior experience on trust: prior linkages bonded in the hard experience of perestroika tended to survive and induce more trust than more recent and more ad hoc associations. A second is the apparent ability of microfinance to increase trust by indirect channels, notably through the making of contacts within government and other agencies. (But some expected indirect routes, such as media exposure, turn out to be irrelevant.) A third is the apparent ability of microfinance to reduce corruption: in Slovakia, the client group experienced significantly less corruption than the control group, particularly in terms of judicial procedure. A fourth factor is the ability of some microfinance groups to reproduce themselves and extend social capital into related areas. In encouraging this process, leadership and intra-group inequality appear to be important; sometimes, as in Poltar, Slovakia, the leadership comes from outside the group. Thus, microfinance does appear to help create social capital in Russia and eastern Europe, but not homogeneously and through some unexpected channels. The following policy implications would appear to emerge: (i) in regions where corruption is serious e.g. Borisoglebsk, MF is particularly valued for its ability to provide an bribe-free source of credit. In so doing, it creates trust which lowers the rate of tax on business transactions—and thence the cost structure of small enterprise. (ii) one of the most important indirect effects of microfinance appears to be its ability to induce greater trust in government officials. Exactly how this effect materialises is not clear, but what is certain is the potential complementarity between NGO microfinance organizations and state organizations in the construction of social capital. (iii) microfinance interventions need to try and maximize the ‘social multiplier’ effects of group organization. One particular channel by which this appears to be possible is by reducing intragroup inequalities. (iv) Trust is the obverse of interpersonal risk, and in those cases where trust is low complementary devices such as insurance supplied by the microfinance provider may help reduce the risk of default. Copyright © 2004 John Wiley & Sons, Ltd.