Introduction. In the Netherlands an ambitious scrapie control program was started at the national level in 1998, based on genetic selection of animals for breeding. From 2002 onwards EU regulations required intensive active scrapie surveillance as well as certain control measures in affected flocks. Materials and Methods. Here we use standard statistical methods as well as mathematical modeling to analyze (1) data on genotype frequencies and scrapie prevalence in the Dutch sheep population obtained from both surveillance and affected flocks; (2) data on PrP genotype frequencies in a random sample of flocks; (3) postal survey results on between-flock differences in breeding strategy and flock management. Results and Conclusions. Analyzing the data (1) we find that the breeding program has produced a steady increase in the level of genetic scrapie resistance in the Dutch sheep population. We also found that a few years later this was followed by a sharp decline in the prevalence of classical scrapie in tested animals. Notably, the estimated classical scrapie prevalence level per head of susceptible genotype declined significantly as well. This indicates that selective breeding has a disproportionate effect on infection prevalence, reminiscent of the well-known population effect of vaccination against a transmissible disease. The overall recent decline in classical scrapie prevalence in Dutch sheep suggests that eradication of the disease in The Netherlands may be within reach. However, a subset of farms may still continue to act as a core group for scrapie transmission for some time, as we show by analyzing between-flock heterogeneities using the data (2) and (3). In addition, genetic resistance levels may decline again in future as participation to the selective breeding program has recently become voluntary.