In a strain of E. coli deficient in RNase III (ABL1), 23S rRNA has been shown to be present in incompletely processed form with extra nucleotides at both the 5' and 3' ends (King et al., 1984, Proc. Natl. Acad. Sci. U.S. 81, 185-188). RNA molecules with four different termini at the 5' end are observed in vivo, and are all found in polysomes. The shortest of these ("C3") is four nucleotides shorter than the accepted mature terminus. In growing cells of both wild-type and mutant strains up to 10% of the 23S rRNA chains contain the 5' C3 terminus. In stationary phase cells, the proportion of C3 termini remains the same in the wild-type cells; but C3 becomes the dominant terminus in the mutant. Species C3 is also one of the 5' termini of 23S rRNA generated in vitro from larger precursors by the action of purified RNase III. We therefore suggest that some form of RNase III may still exist in the mutant; and since no cleavage is detectable at any other RNase III-specific site, the remaining enzyme would have a particular affinity for the C3 cleavage site, especially in stationary phase cells. We raise the question whether the C3 terminus has a special role in cellular metabolism.