At sequential steps of slaughter (scalding, dehairing, singeing, polishing, trimming, washing, and chilling), 200 pig carcasses from two abattoirs were examined for total viable bacteria count (TVC) and the presence of Enterobacteriaceae and coagulase-positive Staphylococcus (CPS) by the wet-dry double-swab technique at the neck, belly, back, and ham. Before scalding, mean TVCs ranged from 5.0 to 6.0 log CFU cm(-2), and Enterobacteriaceae and CPS were detected on all carcasses. At abattoir A, mean TVCs and the percentage of Enterobacteriaceae-positive carcasses were reduced (P < 0.05) after scalding (1.9 log CFU cm(-2) and 12%, respectively), singeing (1.9 log CFU cm(-2) and 66%, respectively), and blast chilling (2.3 log CFU cm(-2) and 17%, respectively) and increased (P < 0.05) after dehairing (3.4 log CFU cm(-2) and 100%, respectively) and polishing (2.9 log CFU cm(-2)). The proportion of CPS-positive samples decreased to < or = 10% after scalding and remained at this level. At abattoir B, mean TVCs and the percentages of Enterobacteriaceae- and CPS-positive carcasses were reduced (P < 0.05) after scalding (2.4 log CFU cm(-2) and 29 and 20%, respectively), polishing (3.7 log CFU cm(-2)), and chilling (2.6 log CFU cm(-2) and 55 and 77%, respectively) and increased (P < 0.05) after the combined dehairing-singeing (4.7 log CFU cm(-2) and 97 and 100%, respectively). Among sites, the neck tended to yield higher levels of contamination from trimming to chilling at both abattoirs (P < 0.05). Consequently, scalding, singeing, and chilling may be integrated in a hazard analysis critical control point (HACCP) system for pig slaughter. As indicated by the higher levels of contamination on carcasses after dehairing-singeing and the following stages at abattoir B, each abattoir should develop its own baseline data and should customize HACCP systems to match process- and site-specific circumstances.