Structural fumigations of food processing plants to manage stored-product insects has been a major component of pest management programs, but limited information on field efficacy is available. Efficacy, based on pheromone trapping data, consists of initial reduction in captures after treatment and rebound in trap captures over time. Pattern of Tribolium castaneum (Herbst) rebound was evaluated after 21 fumigations in two flour mills. Rebound in mean number of beetles captured and the probability of a trap capturing one or more beetles was evaluated. Rebound to a threshold mean beetle capture of 2.5 beetles per trap per 2-wk period took 174 ± 33 d and rebound took longer after fall (248 ± 50 d) than spring (104 ± 21 d) fumigations. Rebound to the probability of capture threshold of 0.50 was 120 ± 21 d, but there was no significant effect of season. Improvement in integrated pest management (IPM) practices in one of the mills was associated with an increase in time to reach mean beetle capture threshold (49 ± 15 d before and 246 ± 71 d after) but not in time to reach the probability of capture threshold (38 ± 14 d before and 165 ± 46 d after). There was a negative correlation between number captured after fumigation and time to rebound to threshold. After improved IPM there was a significant reduction in the number of beetles per trap immediately after fumigation. Above these two thresholds the degree of change in trap captures is significantly greater than below, which suggests they might be useful in evaluating risk in a pest management program.