The immune response after vaccination with infectious bronchitis virus (IBV) under field conditions was measured by the ELISA, haemagglutination-inhibition (HI) and agar-gel precipitin (AGP), tests. Vaccinations were performed in three flocks and one experimental group via the drinking water with the vaccine strains H 120 and H 52. In each flock 40 random serum samples were taken every 2 weeks and tested individually. In the experimental group blood samples were collected every week from each of the 10 chickens. The primary vaccination with H 120 resulted in a rapid increase of antibody titre as detected by ELISA followed by a slow decrease over the next few weeks. By the HI and AGP tests no antibody responses could be seen after this primary vaccination. Revaccination with the H 52 strain provoked a further increase in ELISA titres. In the experimental group, and in flock W, a similar increase occurred by the HI test and precipitating antibodies appeared. The formation of HI antibodies in flock T (nipple waterers) was somewhat retarded and precipitating antibodies were just detectable. In flock F revaccination did not result in the immediate production of HI and AGP antibodies. However, 6 weeks after revaccination a significant rise in ELISA, HI and AGP antibodies was observed, probably as the result of a field infection. It was demonstrated that, based on the higher sensitivity, the ELISA test is more suitable than HI and AGP to monitor antibody responses to vaccination against infectious bronchitis. Strain specificity in the HI test is discussed as a reason for its failure to detect antibodies after primary vaccination with the highly attenuated vaccine strain H 120.