A cross-sectional study was carried out in spring 2007, at the end of the first bluetongue outbreak season, to determine the geographical spread of bluetongue virus serotype 8 (<scp>btv-8</scp>) infection in cattle in the Netherlands and the consequences for some production parameters. Blood samples from cattle submitted to the laboratory of the Dutch Animal Health Service for other voluntary and obligatory health programmes were tested serologically for <scp>btv-8</scp>. In total, 37,073 samples were tested and 659 (1·78 per cent) were seropositive. The samples came from 5436 herds, of which 45 per cent of herds had only one sample submitted from them. The prevalence was highest in the south of the country, where the outbreak had started, and decreased towards the north. In 340 herds more than 50 per cent of cattle were tested, of which 156 herds were located in infected compartments, and in 37 of these herds (10·9 per cent) at least one positive cow was detected. The average within-herd prevalence in the 37 herds was 39·3 per cent: 2·2 per cent in 11 dairy herds, 68·4 per cent in 20 small-scale herds and 14 per cent in four suckler cow herds. The prevalence differed significantly between herd types but did not show a geographical trend. The average net return for milk production amounted to €2417/cow/year and it decreased significantly on average by €48/cow/year in the bluetongue-infected dairy herds during the bluetongue period. On the small-scale farms, the incidence of mortality increased by 3·2 (95 per cent confidence interval [<scp>ci</scp>] 1·2 to 9·1) times in the infected herds during the bluetongue period, but the voluntary culling rate decreased by a factor of 2·3 (95 per cent <scp>ci</scp> 1·1 to 4·8).