• Virus isolation is possible in a wide range of cell cultures; primary bovine kidney, turbinate and testis cells are highly sensitive. The harvesting of the buffy coat from whole blood provides a useful specimen in the live animal. Suitable specimens from post mortem cases include spleen, liver, kidney, lymph node and sections of the gastro-intestinal tract containing lesions. Two samples taken three weeks apart should be analysed to confirm persistent infection in an animal. Cell lines and foetal calf serum, used as a cell culture medium supplement, must be screened for the presence of BVDV before attempting isolation of the virus (Bolin et al. 1991). The foetal calf serum should also be free from BVDV-specific antibodies. Cytopathic BVDV isolates will produce foci of lysed cells in the cell monolayer while non-cytopathic virus can be detected using fluorochrome or enzyme-labelled BVDV-specific antibodies. A microtitre virus isolation method using sera, the immunoperoxidase monolayer assay (IPMA), for whole herd screening for PI cattle has been developed (Saliki et al. 1997). It has been reported that some PI cattle can be IPMA-negative on serum but virus-isolation-positive on buffy coat (Grooms et al. 2001).

• Viral antigen can be detected directly by immunofluorescence on frozen sections or buffy coat smears. Immunohistochemical staining of skin sections (‘ear notch’ test) has been shown to be reliable and to correlate well with findings from blood testing (Thür et al. 1996, Njaa et al. 2000). Several antigen capture ELISAs have been developed for rapid detection of BVDV antigens extracted from tissues or blood leukocytes (Fenton et al. 1991, Shannon et al. 1991) and commercial kit sets are available. These assays are usually based on the use of monoclonal antibodies that recognize the antigenically conserved non-structural protein NS2-3 and therefore should detect most if not all BVDV strains.

• Dot blot, in situ hybridization and RT-PCR techniques for the detection of viral RNA have been described (Roberts et al. 1991, Sandvik et al. 1997, McGoldrick et al. 1999, Letellier & Kerkhofs 2003) as well as a multiplex PCR for the detection and differentiation of BVDV 1 and BVDV 2 (Gilbert et al. 1999). Reverse transcription-PCR is sufficiently sensitive for the detection of viral nucleic acid in pooled samples such as bulk milk (Radwan et al. 1995, Drew et al. 1999) and serum (Weinstock et al. 2001). This is a particularly useful approach where the herd in question has been vaccinated and antibody testing is not appropriate. It is possible to discriminate between BVDV-1 and -2 using separate sets of primers and probes (Letellier & Kerkhofs 2003).

• Virus neutralization and ELISA are the preferred methods for the detection of antibodies to BVDV (Edwards 1990). The virus neutralization test detects antibodies to the viral glycoproteins, particularly E2 (gp55). Although these antibodies are capable of cross-reacting with several strains of BVDV, it is important to use a challenge virus antigenically similar to the field virus circulating in the test population. The tissue culture cells used can also influence the apparent titre of a test serum (Saliki & Dubovi 2004). ELISA kits for detection of BVDV-specific antibodies are available commercially. The collection of paired serum samples and the detection of a fourfold increase in antibody titre are necessary to detect recent infection. The systematic serological testing of bulk milk or pooled blood samples from herds is an important part of identifying herds with PI animals and as such is also a key step in national eradication programmes. In order to confirm the removal of all PI animals from a herd, unvaccinated animals between eight to 12 months should be tested for antibodies. These animals are free of maternal antibody and have been at risk of infection for some time. The use of antibody testing as a screening test for PI animals is not recommended where BVDV vaccines are being used as PI cattle can have neutralizing antibody titres to the vaccine strain of the virus. The NS3 protein, alone or in combination with other BVDV proteins, is frequently used as the basis of commercial ELISAs because it is highly conserved among the pestiviruses. Recent studies have shown that NS3-specific antibody levels in serum and milk are low or undetectable following vaccination with an inactivated BVDV vaccine. As a result it is possible to monitor infection in vaccinated herds using a suitable combination of inactivated vaccine and NS3 antibody test (Makoschey et al. 2007, Kuijk et al. 2008).