In this study, we investigated how the likelihoods of Salmonella presence in various samples from broilers and their grow-out environment throughout one production cycle were related. Sixty-four broiler flocks from 10 complexes of two companies in the southern United States were included in the study. Samples from the gastrointestinal tracts of chicks, transport tray pads and litter and drag swabs from the house were collected on the day of placement of each flock. Approximately, 1 week before harvest, whole bird carcass rinses, caecum and crop samples were collected from birds from these same flocks. On the day of harvest, litter and drag swab samples were also taken from the house after the birds were removed. Upon arrival of the flocks at the processing plant, whole carcass rinses, caecum and crop samples were collected. As the flocks were processed, carcass rinses were collected just before the carcasses entered the immersion chill tank and as they exited the chill tank. Logistic regression was used to model the relationships between the likelihood of Salmonella in samples of each type collected at each sampling point and Salmonella frequencies in all the samples taken from the flock and grow-out environment at preceding production stages. The analysis demonstrated that increased likelihood of Salmonella contaminated carcasses entering the immersion chill tank was associated with higher contamination of the exteriors and crops of birds at arrival for processing as well as house environmental samples at the time of harvest and prior to placement. The best predictors of post-chill broiler carcass Salmonella status were the frequencies of Salmonella in the litter on the day of harvest and prior to placement. The immersion chilling appeared to disrupt some of the relationships between the processing plant and pre-harvest samples.