In Belgium, pseudorabies in swine has been the subject of a mandatory eradication programme since 1993. From December 1995 to February 1996, a survey was conducted in the five provinces of northern Belgium to estimate the provincial pseudorabies virus (PRV) herd seroprevalence. Seven hundred and twenty randomly selected herds were included in this survey. To detect recently infected animals, only young sows were sampled. The results show that 44% of these herds had an important number of PRV-seropositive young sows. The highest herd seroprevalence was observed in West Flanders (68%), followed by Antwerp (60%), East Flanders (43%), Limburg (18%), and Flemish Brabant (8%). Assuming a diagnostic test sensitivity and specificity of 95% and 99%, respectively, and a true PRV within-herd prevalence of 43%, the overall true PRV herd prevalence was estimated to be 35%. A logistic multiple-regression revealed that the presence of finishing pigs was associated with a two-fold increase in odds of a herd being seropositive (odds ratio (OR)=2.07, 95% confidence interval (CI) = 1.31-3.26); a breeding herd size > or =70 sows was associated with a four-fold increase in odds of a herd being seropositive (OR = 4.09, 95% CI = 2.18-7.67); a pig density in the municipality of >455 pigs/km2 was associated with a 10-fold increase in odds of a herd being seropositive (OR = 9.68, 95% CI = 5.17-18.12). No association was detected between the PRV herd seroprevalence and purchase policy of breeding pigs (purchased gilts, or use of homebred gilts only).