Measurement of feline coronavirus (FCoV) antibody titres is utilised mainly for diagnosing feline infectious peritonitis (FIP) and for quarantine purposes. However, occasional samples show a falsely low or negative FCoV antibody test. We tested the hypothesis that such results are due to virus in the sample binding antibody and rendering it unavailable to antigen in the test. Thirteen effusions, one plasma and three undefined samples from cats with FIP, which gave unexpectedly low FCoV antibody titres, were examined by real-time reverse transcriptase polymerase chain reaction (RT-PCR). Increasing amounts of virus correlated with lower signals in indirect immunoflourescent, enzyme-linked immunosorbent asssay and rapid immunomigration antibody tests. However, five samples were negative by RT-PCR, so the presence of virus alone may not explain all cases of false-negative FCoV antibody tests, although it is a possible explanation in 71% of discordant samples. We conclude that falsely low or negative FCoV antibody tests can occur in samples rich in virus.