A recurrent theme in ethnomethodological research is that of instructed actions. Contrary to the classic traditions in the social and cognitive sciences, which attribute logical priority or causal primacy to instructions, rules, and structures of action, ethnomethodologists investigate the situated production of actions which enable such formulations to stand as adequate accounts. Consequently, a recitation of formal structures can not count as an adequate sociological description, when no account is given of the local production ofwhat those structures describe. The natural (...) sciences can be described as a domain of practical action in whichthe use of methods enables the intersubjective reproduction of naturalistic observations and experiments. As numerous sociological studies of laboratory practices have shown, the achievement of intersubjective order cannot be reduced to formal methods; instead, it arises from the work of custom-fitting relevant methods to the local circumstances of the research. In this paper we discuss a possible extension of this idea to cover two intertwined aspects of molecular biology: (1) the work of following instructions on how to perform routine laboratory procedures, and (2) the relationship between cellular orders and the encoded instructions contained in the DNA molecule. We suggest that a classic conception of scientific action is implied by the way formal instructions are treated as a primary basis, both for molecular biologists' actions and the cellular functions they study, and we envision an ethnomethodological alternative to those conceptions of social and biological order. (shrink)