Modern biology is rapidly laying the foundation necessary for integrated modeling of physiological processes in living organisms. The human physiome project attempts to model interactions between biochemicals, cellular organelles, cells, tissues, and organs within whole organisms. One of the first challenges that this project faces is the development of a database environment flexible enough to accommodate the diversity in structure and content of physiological data. This paper reviews the current state of database technology, presents our understanding of the physiome database problem, and proposes a preliminary strategy for addressing it.