The aim of this study was to integrate an ozone-based sanitization step into existing processing practices for fresh produce and to evaluate the efficacy of this step against Escherichia coli O157:H7. Baby spinach inoculated with E. coli O157:H7 (approximately 10(7) CFU/g) was treated in a pilot-scale system with combinations of vacuum cooling and sanitizing levels of ozone gas (SanVac). The contribution of process variables (ozone concentration, pressure, and treatment time) to lethality was investigated using response-surface methodology. SanVac processes decreased E. coli O157:H7 populations by up to 2.4 log CFU/g. An optimized SanVac process that inactivated 1.8 log CFU/g with no apparent damage to the quality of the spinach had the following parameters: O3 at 1.5 g/kg gas-mix (935 ppm, vol/vol), 10 psig of holding pressure, and 30 min of holding time. In a separate set of experiments, refrigerated spinach was treated with low ozone levels (8 to 16 mg/kg; 5 to 10 ppm, vol/vol) for up to 3 days in a system that simulated sanitization during transportation (SanTrans). The treatment decreased E. coli populations by up to 1.4 log CFU/g, and the optimum process resulted in a 1.0-log inactivation with minimal effect on product quality. In a third group of experiments, freshly harvested unprocessed spinach was inoculated with E. coli O157:H7 and sequentially subjected to optimized SanVac and SanTrans processes. This double treatment inactivated 4.1 to > or = 5.0 log CFU/g, depending on the treatment time. These novel sanitization approaches were effective in considerably reducing the E. coli O157: H7 populations on spinach and should be relatively easy to integrate into existing fresh produce processes and practices.