In Escherichia coli cultures limited for phosphate, the number of ribosomal particles was reduced to a small percentage of its earlier peak value by the time the viable cell count began to drop; the 30S subunits decreased more than the 50S subunits. Moreover, the ribosomal activity was reduced even more: these cells no longer synthesized protein, and their extracts could not translate phage RNA unless ribosomes were added. The translation initiation factors also disappeared, suggesting that they become less stable when released from their normal attachment to 30S subunits. In contrast, elongation factors, aminoacyl-tRNA synthetases, and tRNA persisted. During further incubation, until viability was reduced to 10-5, the ribosomal particles disappeared altogether, while tRNA continued to be preserved. These results suggest that an excessive loss of ribosomes (and of initiation factors) may be a major cause of cell death during prolonged phosphate starvation.Formidable methodological obstacles have discouraged investigation of the mechanism of the loss of viability of bacteria in the stationary phase (25), although the frequent exposure of bacteria to prolonged starvation in nature makes this an important problem. In approaching it we suggest that the theoretically predictable mechanisms of cell death fall into only three classes. The first mechanism is irreversible damage to a vital element in the genome. This seems unlikely to be a major response to starvation (unless a prophage is activated), since cells have elaborate devices for protecting the DNA. The second mechanism is irreversible damage to the cell membrane. Although this is clearly important in organisms that autolyze readily during starvation (such as Streptococcus pneumoniae), enterobacteria and many other organisms maintain their turbidity. Even stronger evidence against this mechanism is the observation of Postgate and Hunter (26) that starved Aerobacter aerogenes cells maintained an intact osmotic barrier after the viability count had dropped by 80%.The third mechanism is the complete loss of certain species of macromolecules. These would not include most enzymes, the cofactors, any kind of RNA, or the components of various energy-transducing systems, since these products should all be restorable if their genes are intact and if the necessary building blocks and energy can be provided. However, the complete loss of any protein species required for protein synthesis should be lethal. Indeed, this mechanism has been demonstrated: temperature inactivation of a temperature-sensitive mutant elongation factor G rapidly destroyed viability (1). The same effect would be expected from the complete loss of ribosomes, of any soluble factors of translation or aminoacyl-tRNA synthetases, or of RNA polymerase.The degradation of ribosomes seems a likely mechanism in this third class, since it is a well-known major response of bacteria to starvation (17). This adaptive process provides nucleotides and amino acids, and it might also provide energy in the form of nuc...