Overproduction of rRNA was artificially induced in Escherichia coli cells to test whether the synthesis of ribosomal protein (r-protein) is normally repressed by feedback regulation. When rRNA was overproduced more than twofold from a hybrid plasmid carrying the rrnB operon fused to the lambda pL promoter (pL-rrnB), synthesis of individual r-proteins increased by an average of about 60%. This demonstrates that the synthesis of r-proteins is repressed under normal conditions. The increase of r-protein production, however, for unknown reasons, was not as great as the increase in rRNA synthesis and resulted in an imbalance between the amounts of rRNA and r-protein synthesis. Therefore, only a small (less than 20%) increase in the synthesis of complete 30S and 50S ribosome subunits was detected, and a considerable fraction of the excess rRNA was degraded. Lack of complete cooperativity in the assembly of ribosome subunits in vivo is discussed as a possible explanation for the absence of a large stimulation of ribosome synthesis observed under these conditions. In addition to the induction of intact rRNA overproduction from the pL-rrnB operon, the effects of unbalanced overproduction of each of the two large rRNAs, 16S rRNA and 23S rRNA, on r-protein synthesis were examined using pL-rrnB derivatives carrying a large deletion in either the 23S rRNA gene or the 16S rRNA gene. Operon-specific derepression after 23S or 16S rRNA overproduction correlated with the overproduction of rRNA containing the target site for the operon-specific repressor r-protein. These results are discussed to explain the apparent coupling of the assembly of one ribosomal subunit with that of the other which was observed in earlier studies on conditionally lethal mutants with defects in ribosome assembly.