The World Health Organization has declared the scalding, defeathering, evisceration, and chilling processing stages in industrial poultry slaughterhouses as sites of cross-contamination. In response, we develop a novel ordinary differential equation model describing the transfer and thermal inactivation of pathogens between carcasses in an industrial three-stage scalding tank. The model provides a link between the distribution of bacteria on carcasses before and after scalding using physical mechanisms of transfer and inactivation. A key feature of our model is that it directly quantifies cross-contamination, or microbial transfer between carcasses, via the scald water. We parametrize the model according to the specific pathogen C. jejuni, and then validate the model using numerical simulations against industrial scalding processing data. In particular, the calibrated model predicts a mean pathogen level on individual carcasses of [1.15 x 10(4), 2.6 x 10(4)] CFU/carcass, capturing the experimentally determined mean of 1.89 x 10(4) CFU/carcass in an industrial setting. We then demonstrate that our framework is amenable to application in quantitative microbial risk assessment and policymaking. In particular, using scalding configurations from an industrial poultry processor in Canada as well as the 2015 USDA-FSIS Campylobacter performance standards as a benchmark, we utilize the developed model to quantify the relationship between pre-scald prevalence and mean pathogen counts and the distribution of post-scald pathogen levels on carcasses.