Low temperatures are widely used to ensure food quality and safety. However, sublethally injured Staphylococcus aureus is an important microbiological safety concern in low temperature food. The objective of this study was to develop predictive inactivation kinetic models for the inactivation and sublethal injury of S. aureus in broth at different temperatures (4 to -18°C) and time points. S. aureus was diluted in tryptic soy broth plus 0.6% (wt/vol) yeast extract (TSBYE) to obtain approximately 10(8) CFU/ml and was stored separately at 4, -3, -11, and -18°C. After specific time points within 96 days, survival of S. aureus was determined on TSBYE and TSBYE agar plus 10% NaCl for enumeration of the total viable and noninjured cell numbers, respectively. Linear, Weibull, and modified Gompertz models were applied to determine survival curve regression. The combination of low temperature and time resulted in S. aureus inactivation, although the cells were able to survive in this sublethal state. Storage temperature was the critical parameter in survival of S. aureus. The modified Weibull model successfully described a second model of noninjured S. aureus cell survival at different low temperatures, whereas only the linear model was able to fit the total viable cells. The predictive model may be used to estimate the level of S. aureus contamination in food at low storage temperatures and times, and it provides new insight into the sublethally injured survival state of S. aureus in low temperature food.