The technique of continuous culture has been used to study the segregatlonal stability of plasmids in Escherichia coli in the absence of selective pressure. Conditions were established that allowed the detection of plasmid-free cells no matter how low their Initial frequency. Using these conditions the segregational stability of two related multicopy plasmids (pDSII09 and pBR322) was examined. pDSII09 was found to be stably Inherited throughout 120 generations of nutrient limited growth despite the observation that the plasmid copy number fell 4 to 5 fold during the culture period. By contrast, pBR322 was lost from chemostat culture after a lag of between 30 and 40 generations, a period during which (by analogy with pDSII09) its copy number had fallen about 2 fold. The functional basis of the differential segregation exhibited by these plasmids was ascribed to the presence (on pDSII09) or absence (on pBR322) of a functional par (partition) signal that ensured the efficient segregation of plasmid molecules into daughter cells at division. Based on this hypothesis, experiments were done to examine the possibility for correction of the defective partitioning of pBR322 by complementation in cis and in trans. In only two cases (both in cis) was complementation achieved. The first using a previously characterised par function from plasmid pSCIOI and the second using a fragment of plasmid pDSII09 that was considered (by argument) to be the region involved in its observed segregational stability. Supportive evidence for the existence of partition elements amongst multicopy plasmids is cited, as is confirmatory work since published by other workers. A possible mechanism of par action is discussed. Chemostat culture has also been used to examine the possibility of plasmid transfer by transformation within the chemostat. This study examined the effects of both growth rate and nutrient limitation on the transformability of Escherichia coli grown in continuous culture. The results obtained are discussed in relation to previously published transformation work using batch grown cells and a possible mechanism of plasmid transformation is suggested.