We have confirmed the report (Mangiarotti & Schlessinger, 1966) that a new method of lysing Escherichia coli cells yields polysomes and ribosomal subunits but no 70 s ribosomes. However, two features of that method are shown to promote ribosome dissociation: growth in hypertonic medium, and the use of an excess of deoxycholate (which precipitates Mg2+). When the cells are lysed by a freeze-thaw-lysozyme method, with a lower deoxycholate concentration, a substantial 70 s peak is seen along with polysomes and subunits. Moreover, when the products of polysome run-off are caused to accumulate in the cells by treatment with actinomycin D or with puromycin, or by starvation for the carbon source or for a required amino acid, they appear only as 70 s particles. Finally, treatment of complexed 70 s particles (fragmented polysomes) with puromycin in vitro does not cause dissociation into subunits, though the nascent polypeptide is removed. We conclude that (a) polysomes release their run-off ribosomes as stable, free 70 s units, and (b) the dissociation of these units in the cell is subject to a physiological control which maintains an essentially constant pool of subunits. These findings, together with recent work on initiation factors, suggest that dissociation of ribosomes depends on complexing with a factor that is present in the cell in limited supply. © 1968.