The digestion of heat-killed/stained Escherichia coli, Pseudomonas aeruginosa, Mesorhizobium sp. and Staphylococcus aureus was monitored within one food vacuole passage time in the ciliate Tetrahymena pyriformis using the pulse chase technique. Prey digestion proceeded in 2 phases: a digestive phase which lasted ~25 min and a defecation competent phase that showed limited digestive activity and was variable in length. The number of prey cells per food vacuole was found to influence the effectiveness of prey digestion. Complete digestion of the vacuole content was more likely to occur when the number of prey per vacuole averaged ~6 or less. At higher levels, only partial digestion of the vacuole content was recorded and some undigested prey were egested from the ciliate cell. A strain of Synechococcus sp. was never digested by this ciliate. Results suggest that bacteria do not necessarily require elaborate mechanisms to evade digestion by protozoa, as possessed by some pathogenic bacteria, but that inefficiency in the protozoan digestive system is all that is required to allow the release of undigested, apparently unharmed prey from their cells. Thus, models on carbon cycling which employ data on protistan ingestion rates alone should consider accounting for digestion efficiency and the subsequent effect of prey concentration, because prey carbon might not always be transferred efficiently to higher trophic levels.