The growth of bacteria in pure cultures has been the mainstay of microbiologi cal technique from the time of Pasteur to the present. Solid media techniques have allowed the isolation of individual species from complex natural pop ulations. These pure isolates are intensively studied as they grow in batch cultures in nutrient-rich media. This experimental approach has served well in providing an increasingly accurate understanding of prokaryotic genetics and metabolism and in facilitating the isolation and identification of pathogens in a wide variety of diseases. Further, vaccines and antibiotics developed on the basis of in vitro data and tested on test-tube bacteria have provided a large measure of control of these pathogenic organisms. During the last two decades microbial ecologists have developed a series of exciting new techniques for the examination of bacteria growing in vivo, and often in situ, in natural environments and in pathogenic relationships with tissues. The data suggest that these organisms differ profoundly from cells of the same species grown in vitro. Brown & Williams (12) have shown that bacteria growing in infected tissues produce cell surface components not found on cells grown in vitro and that a whole spectrum of cell wall structures may be produced in cells of the same species in response to variations in nutrient status, surface growth, and other environmental factors (6 7). We and others (28) have used direct ecological methods to examine bacterial cells growing in natural and pathogenic ecosystems, and we find that many impor tant populations grow in adherent biofilms and structured consortia that are not seen in pure cultures growing in nutrient-rich media. In fact, it is difficult to imagine actual natural or pathogenic ecosystems in which the bacteria would be as well nourished and as well protected as they are in single-species batch cultures.