A serological survey was carried out to detect specific (serotype 20) and a group bluetongue virus antibody in cattle and sheep serums collected in Western Australia during the period January 1 1978 to June 30 1979. Of 18,849 cattle serums examined by the gel diffusion precipitin test (GDPT), 9.7% were positive and 6.1% gave doubtful results. All 1949 sheep serums tested were negative. Precipitin antibody was demonstrated in 22.5% of serums from Kimberley cattle and 3.6% of cattle serums from the Northwest. Serums collected from cattle in the South were consistently negative in GDPT. When 915 serums that reacted in the GDPT were further tested by the complement fixation test (CFT), 164 were positive. The percentage of CFT positive serums increased as the GDPT reaction became stronger. 2467 serums collected from cattle in Kimberley and Northwest areas and tested by the CFT, 175 (7.1%) were positive. These 175 positive serums were also examined by GDPT and 164 doubtful or positive reactions were obtained. The virus neutralisation (VNT) using serotype 20 virus was carried out on 3804 serums, including all serums that reacted in the GDPT, and 57 were positive. When the VNT positive serums were examined in the other 2 tests, 47 serums were either positive or doubtful in the GDPT and 8 were positive in the CFT. The presence of bluetongue virus group antibody in cattle serums closely followed the suggested distribution pattern of Culicoides brevitarsis but specific serotype 20 neutralising antibody was limited to cattle serums from stations situated north of latitude 17 degrees S in an area of mean annual rainfall higher than 700 mm.