Members of the genus Salmonella represent a significant public health concern and also a colonizer of commercial poultry. Therefore, the early detection and management of colonized broiler breeders and their progeny is essential. There have been numerous methods for farm-based detection, with gauze-based drag swabs being the most commonly used. In the present study, the wet (boiled water, buffered peptone water and double-strength skin milk) tampon was compared with the gauze to determine the recovery rate (10(2) colony-forming units/swab) of five common poultry serovars of Salmonella and after cold (4°C/48 h) storage. The recovery was found to be equivalent when tested using the ISO6572:2002 method, for all diluents (Cohen's κ =1.0; sensitivity = 1.0; specificity = 1.0). The subsequent field trial (n = 15 farms) compared the tampon drag swab (TDS) with a statistically appropriate (90% confidence, detect 10% prevalence) number of faecal swabs (n = 22), which also showed high agreement between the TDS and faecal sampling (κ = 0.86; McNemar's χ(2) = 1.0; sensitivity = 0.9; specificity = 1.0). However, direct faecal sampling showed a wider diversity of serovars of Salmonella than the corresponding TDS. The TDS is a very sensitive, readily available and cost-effective screening method for salmonellas in broiler breeder houses. This TDS technique may be used for routinely screening of broiler houses, and faecal sampling would only be used to confirm colonization or contamination, and to measure flock serovar variance.