A plasmid vector pBK2 was tested for maintenance in Corynebacterium acetoacidophilum and found to be 100% stable for 90 generations. In co-culture experiments in absence of any selection pressure a was able to cause washout of plasmid free cells. Deletion mutagenesis of the plasmid was done to identify the regions of DNA responsible for its stable maintenance. This led to the identification of a region of DNA in pBB1, an endogenous plasmid used in the construction of pBK2, which is responsible for providing a growth rate advantage to the plasmid containing cell in co-culture with plasmid;free cells. To explain this observed stability we postulate the production of a "growth-inhibiting factor" by the plasmid-containing cells. Inclusion of a death rate term for the plasmid free cells in the computer simulation of the growth kinetics in a chemostat successfully predicts, depending upon the magnitude of the kinetic parameters, a stable coexistence of the two or washout of the plasmid-free cells, an observation that matches the experimental data. Because of its stable maintenance and its ability to select against plasmid-free cells this host-vector system is a natural candidate for cloning and use in large-scale fermentations.