This paper describes the validation of a solid-phase competition enzyme-linked immunosorbent assay (SPCE) for the serological detection of antibody to serotype O foot-and-mouth disease (FMD) in sheep, cattle and pigs. The specificity of the SPCE was calculated from the results of testing known negative sera from sheep, cattle and pigs (n=3030, 1418 and 1495, respectively). The mean percentage inhibition (PI) for known negative sheep, cattle and pig sera were 19.3, 24.1 and 20.8%, respectively. The specificity of the SPCE at a cut-off point (COP) of 60 PI was 99.50% for sheep sera, 99.44% for cattle sera and 100% for pig sera. The analytical sensitivity of the SPCE was examined by testing sera from sheep, cattle and pigs. Based on the testing of serial bleeds from experimentally infected animals, seroconversion at the 60 PI COP occurred between 4 and 9 days post-infection or -exposure, similar to that observed using the virus neutralisation test (VNT) with a COP of 1/45. When applied to 267 sheep and 143 pig samples, that were obtained in Great Britain (GB) during the 2001 FMD UK outbreak, the SPCE identified more positive samples than did the VNT. Estimates of the accuracy, repeatability and reproducibility of the SPCE were verified during the large-scale serosurveillance necessitated by the 2001 outbreak. Results from field and experimental sera showed that when compared against the VNT, the sensitivity of the SPCE was less affected by the choice of virus strain used in the test. Using the O(1) UKG 2001 FMD virus in the VNT with samples representative of the uninfected GB sheep population, the test specificity was 100% at a COP of 1/45.