Escherichia coli ATCC 25922 in phosphate buffered saline solution (PBS, pH 7.1, 10(8) CFU/ml) was inactivated by high hydrostatic pressure (HHP, 400 to 600 MPa) treatment at 25 degrees C for 10 min. Colonies of E. coli were not detected on non-selective plate count agar immediately after a HHP-treatment of at least 550 MPa. E. coli subjected to at least 500 MPa in PBS were incubated at 4, 25, and 37 degrees C for 120 h. No colonies were detected on plate count agar throughout the 120-h incubation period at 4 or 37 degrees C. In contrast, the number of E. coli during storage at 25 degrees C increased from an undetectable level (< 1 CFU/ml) to the level of initial cell counts regardless of the treatment pressure level. The recovery in PBS required a maximum time of 48 h, while the period during which cell numbers remained at an undetectable level increased from 24 to 72 h as the treatment pressure increased. E. coli treated at 550 and 600 MPa in PBS were inoculated into trypticase soy broth (TSB) and stored at 4, 25, and 37 degrees C for 120 h. No recovery was recorded in TSB during the 120-h storage at 37 degrees C. In contrast, the number of E.coli during storage at 25 degrees C in TSB increased beyond the level of initial cell counts regardless of the treatment pressure level. The recovery of cell numbers observed in TSB was faster than that in PBS samples, as bacterial growth in TSB assisted faster recovery. When the incubation temperature in PBS was shifted to 25 degrees C after 120-h at 4 or 37 degrees C, recovery of E. coli was observed in samples shifted from 4 to 25 degrees C regardless of the treatment pressure. However, the time during which cell numbers remained at an undetectable level was extended by increasing the level of treatment pressure, and recovery required a maximum time of 48 h. On the other hand, no recovery was observed with HHP-treated E. coli subjected to an incubation temperature shift from 37 to 25 degrees C. This study indicates that an appropriate incubation temperature after HHP-treatment is needed to optimize the recovery of HHP-injured bacteria and thus prevent overestimation of the lethal effect of HHP-treatment.