There are three sources of information regarding the social, political, and economic status of women in the European Iron Age, and all three are problematic. First, there are the classical sources, which describe Celtic society from the ‘civilised’ Mediterranean perspective, whether Greek or Roman. These often tell us more about the observer's culture than about the Celtic societies being described. Second, there are the insular Celtic laws and legends, which are compromised by time, distance, and the transformative effects of Christianity. Third, there is the archaeological record, which can only produce answers if the questions asked are the right ones, and is often seriously flawed with respect to preservation and recovery. Unfortunately, archaeologists are still some way from developing a methodology of reconstructing gender roles on the basis of excavated material culture remains. I would like to discuss an archaeological example which may provide some useful guidelines for generating such a methodology, with reference to all three of the sources mentioned above. The approach presented here makes use of written, ethnographic, and archaeological sources in a way that is not always possible in all research contexts. However, there have been a few recent attempts at such a synthesis (Claassen 1992; Spector 1993; Robb 1994), which suggest that a partnership between the written, ethnographic, and archaeological records regarding the subject of gender yields more than any of these sources of information alone can produce. Where such complementary data sets exist they should be made to work in tandem, as this discussion will show.