This study investigates the performance of meat and poultry plant managers in discovering and responding effectively to food safety problems that lead to product recalls. Timing is used as a performance measure of managers' response to recalls of food, using survival distributions of times between production and recall, and recall case duration. The objectives are to understand how these time periods vary across plants and to determine factors explaining such variability. Survival distributions are estimated using the Kaplan-Meier and life table methods. Subgroups of the population are compared using plots of the estimated survival functions and statistically compared using log-rank and Wilcoxon tests. Managers at large plants, in multi-plant firms, and at plants with prior recall experience do not perform better. Cox regressions indicate that government agency sampling programs enhanced the speed of discovery, and that national distribution networks contributed to the risk that cases remained open for a longer period. [EconLit citations: D210, Q180.] © 2005 Wiley Periodicals, Inc. Agribusiness 21: 351–373, 2005.