The prevalence of bovine viral diarrhoea (BVD) serologically positive animals in 18 dairy herds with clinical and pathological lesions suggestive of BVD infection, the post-vaccinal seroconversion rates in negative animals vaccinated twice with an inactivated BVD vaccine, and the control measures taken, are described. The pathological and histopathological findings in 6 necropsies performed on animals that died in 5 separate herds closely resembled published descriptions. Positive immunohistochemistry results in 3 cases confirmed the diagnosis in those animals. In 1 herd the prevalence of prevaccinal BVDantibodies was only 36.8 %, while the prevalence varied from 79.85 to 100 % in the remainder. Control measures taken included immunoprophylaxis with an inactivated vaccine, culling animals that were serologically negative after vaccination that were regarded as probably persistently infected (PI) and the implementation of additional biosecurity measures. The prevalence of serologically negative PI animals in 10 herds varied from 0.38 to 4.04 %, with 8 herds less than 1 %and 2 herds at 2.79 %and 4.04 %, respectively. Methods based on vaccinating the herd, followed by serological testing and culling cattle that did not develop an antibody titre, are not reliable. The identification of PI animals should be confirmed by isolation of the virus or identification of the antigen.