Anthropologists writing on the Highlands societies of Papua New Guinea have stressed the variable importance of ideas of menstrual pollution as markers of gender relations. This article suggests an alternative approach to these ideas, emphasizing instead aspects of power, placement, complementarity, collaboration, and the moral agency of both genders. Turning to the ethnographic work of the 1960s, we contrast the writings of Salisbury and Meggitt and discuss the usefulness of the "three bodies" concept of Lock and Scheper-Hughes in the comparative analysis of body substances and their meanings in this region. The use of a collaborative model is helpful in such an overall analysis. (Bodily substances, cosmos, gender, menstruation, power)' Throughout the world menstrual blood is seen as a powerful bodily fluid in many contexts, especially in its recognized role as a source of fertility, and it is a significant part of a cosmic cycle of life and death. Menstrual blood may be used, like other powerful bodily fluids, in beneficial ways or in lethal ways. In interpreting the use of this fluid the agency of both women and men needs to be considered. Likewise, the cosmological contexts in which bodily substances are regarded must be explored so as to provide a better understanding of how these fluids are seen in all their dimensions (Stewart and Strathern 2001a, 1999a). Buckley and Gottlieb (1988:7) discuss the diverse practices encompassed by a category such as menstrual taboos. These taboos affect the behavior of others in addition to the menstruating woman, and these actions have "cosmological ramifications" (Buckley and Gottlieb 1988:9). They say there is a sort of androcentric bias in the interpretation that anthropologists have frequently made of taboos generally. For example, if women are prohibited from touching male hunting gear, this is interpreted as representing male dominance, but if men are not allowed to touch something in the female domain, this is often interpreted as a sign of female inferiority (Buckley and Gottlieb 1988:14). Rather than demonstrating absolute representations of superiority/inferiority, taboos such as those surrounding menstruation are best understood through observation and interpretation of the specific ways that taboos operate within a local cultural logic. This article uses examples from the Highlands of Papua New Guinea to explore this topic. We use ethnographic records of other anthropologists and draw on materials from our three field areas in the Highlands of Papua New Guinea, the Hagen, Duna, and Pangia (Wiru) areas. The Hagen area belongs to the populous Wahgi Valley in the Western Highlands Province, with a population today of around 100,000 and an administrative center in Mount Hagen township. The Pangia area, inhabited by more than 20,000 speakers of the Wiru language, lies to the south of Hagen and belongs to the Southern Highlands Province, whose headquarters is in the 349