The efficacy of a commercial swine influenza vaccine based on A/New Jersey/8/76 (H1N1) and A/Port Chalmers/1/73 (H3N2) strains was tested against challenge with an H1N2 swine influenza virus. Influenza virus-seronegative pigs were vaccinated twice with the vaccine when they were four and eight weeks old, or with the same vaccine supplemented with an H1N2 component. Control pigs were left unvaccinated. Three weeks after the second vaccination, all the pigs were challenged intratracheally with the swine influenza strain Sw/Gent/7625/99 (H1N2). The commercial vaccine induced cross-reactive antibodies to H1N2, as detected by the virus neutralisation (VN) assay, but VN antibody titres were 18 times lower than in the pigs vaccinated with the H1N2-supplemented vaccine. The challenge produced severe respiratory signs in nine of 10 unvaccinated control pigs, which developed high H1N2 virus titres in the lungs 24 and 72 hours after the challenge. Vaccination with the commercial vaccine resulted in milder respiratory signs, but H1N2 virus replication was not prevented. Mean virus titres in the pigs vaccinated with the commercial vaccine were 1-5 log10 lower than in the controls at 24 hours but no different at 72 hours. In contrast, the H1N2-supplemented vaccine prevented respiratory disease in most pigs. There was a 4-5 log10 reduction in the mean virus titre at 24 hours in the pigs vaccinated with this vaccine, and no detectable virus replication at 72 hours. These data indicate that the commercial swine influenza vaccine did not confer adequate protection against the H1N2 subtype.