At first view, the methodology of survey research may seem rather unsuitable to the study of such "holistic" phenomena as collective and social identities. That difficulty vanishes - at least partly - as soon as we consider social identity as the link between the individual and his belongings, as does the "social identity theory", developed from the work of Taffel and Turner. From there on, survey research may prove to be a useful device to cope with some main characteristics of social identity: mainly its variability among groups and classes within a same society and its particular sensitivity to socio-political contexts. Survey research, combined with the social identity theory may help to test historical assumptions at a macro-social level. It may also give some ''flesh" and some additional realism to the micro-theories of social behaviour, which are too often limited by their conception of a strictly rational and interested agent.