FOOT-and-mouth disease (FMD) is a contagious viral disease of cloven-hoofed animals including ruminants and pigs. The occurrence of disease in livestock has a great economic impact, especially for exporting countries. Export limitations are based partly on the existence of FMD carrier animals. Carriers have been defined as animals in which virus is present more than 28 days after infection (Sutmoller and Gaggero 1965). Approximately 50 per cent of cattle become carriers of FMD virus (FMDV) after clinical or subclinical infection; this has been observed in both vaccinated and unvaccinated animals. Other ruminants, such as sheep, goats and African buffaloes, have also been recognised as carriers of FMDV. A number of studies have tried to define a carrier state in pigs. Mezencio and others (1999) described the finding of carrier pigs by identifying FMDV RNA in sera from pigs 226 days after infection. However, Alexandersen and others (2003) showed that pigs cleared the virus within three to four weeks. This short communication describes a real-time reverse transcriptase (RT)-PCR study to identify FMDV RNA in the tonsils of pigs 31 to 32 days after initial inoculation with FMDV. The study focused on the tonsil because in ruminants the oropharynx is considered to be an important site of viral persistence. Tonsils were available from two transmission experiments performed with pigs 10 weeks of age, as described by Orsel and others (2007). Twenty-five pigs were vaccinated with a standard dose of double oil emulsion FMD O1Manisa vaccine 15 days before infection. The pigs were infected with FMDV O/NET/2001 either by intradermal inoculation in the bulbus of the heel or by contact with infected pigs. The pigs were slaughtered 30 to 32 days after infection. A biopsy of 1 x 1 cm was collected from the central part of the left tonsil. Tissue suspensions were prepared by disrupting the fresh-frozen biopsy in phosphate-buffered saline containing 2 per cent fetal calf serum and 10 per cent mixed antibiotics using the MagNA Lyser (Roche). All the samples were handled in a class II laminar flow cabinet to prevent contamination. The tissue suspensions were tested by virus isolation on a monolayer of secondary lamb kidney cells. In total, 200 µl of tonsil tissue suspension was added to one well. After one hour's incubation, the wells were washed with fresh medium and 2·5 ml of fresh medium was added. The cells were macroscopically observed for cytopathic effects (cpe) for two days. If no cpe was observed, the cells and supernatant were frozen and thawed, and 200 µl of this suspension was tested in the same way. All incubations were performed at 37°C in a humidified atmosphere containing 5 per cent carbon dioxide. The tonsil suspensions were also tested by automated real-time RT-PCR. RNA isolation was performed as described in detail by Moonen and others (2003). The RT-PCR was performed using the LightCycler system (Roche) with hybridisation probes performed in a closed glass capillary, thereby minimising the risk of cross-contamination. The RT-PCR for FMDV is ISO 17025 accredited and validated within the laboratory facilities of CIDC-Lelystad. No virus was isolated from the tonsil samples from the pigs. In contrast, two of the 25 vaccinated animals and 11 of 31 non-vaccinated control animals tested positive by RT-PCR (Tables 1, 2). The difference between the number of positive samples from the vaccinated and non-vaccinated pigs was significant (P=0·024) Fisher's exact test. All RT-PCR-positive samples originated from pigs that showed clinical signs and from which virus had been isolated from oropharyngeal fluid (OPF) samples collected by swabbing during the acute stage of FMD infection, but from which no virus had been detected in OPF samples for at least the last 22 days of the study period. It was not clear what the biological relevance was of the low amounts of viral RNA, since no viable virus was detected.