Abstract Background In August 2006 a major epidemic of bluetongue virus serotype 8 (BTV8) started off in North-West Europe. In the course of 2007 it became evident that BTV8 had survived the winter in North-West Europe, re-emerged and spread exponentially. Recently, the European Union decided to start vaccination against BTV8. In order to improve the understanding of the epidemiological situation, it was necessary to execute a cross-sectional serological study at the end of the BT vector season. Cattle were the target species for cross-sectional serological studies in Europe at the end of 2006 and 2007. However, there was no information on the BTV8-seroprevalence in sheep and goats. Results On the basis of our cross-sectional study, the estimated seroprevalence of BTV8-exposed locations in the Netherlands in 2006 was 0% for goats (95% confidence interval: 0 – 5.6%) and 7.0% for sheep (95% confidence interval: 3.5 – 12.9%). The estimated seroprevalence of BTV-8 exposed locations in 2007 was 47% for goats (95% confidence interval: 36 – 58%) and 70% for sheep (95% confidence interval: 63 – 76%). There was a wide range in within-location seroprevalence in locations with goats and sheep (1 – 100%). A gradient in seroprevalence was seen, with the highest level of seroprevalence in the southern Netherlands, the area where the epidemic started in 2006, and a decreasing seroprevalence when going in a northern direction. Conclusion There is a much higher estimated seroprevalence of locations with goats exposed to BTV8 than can be inferred from the rather low number of reported clinical outbreaks in goats. This is probably due to the fact that clinical signs in infected goats are far less obvious than in sheep. The wide range in within-location seroprevalence observed means that the proportion of animals protected in 2008 by a natural infection in 2006 and/or 2007 can differ highly between flocks. This should be taken into account when vaccinating animals.