Subclinical infection in scrapie of sheep, characterized by a long incubation period, may be of importance for the spread of the disease. We screened brain samples from all 65 sheep in a scrapie-affected flock for subclinical infection and correlated with results of PrP genotyping, which is of relevance for the epidemiology and the question, whether by breeding for resistant genotypes one would be breeding for healthy carriers. The sensitivity of three methods was compared, i.e. histopathological examination for vacuoles (HP), immunohistochemical staining (IHC) and Western blotting (WB) for PrP(Sc). Five sheep showed definite clinical signs and histological scrapie lesions, and signs of infection were detected in 25 of 60 asymptomatic sheep, by HP and/or IHC and WB. The IHC was slightly more sensitive than HP and WB. Sheep with subclinical infection were, with one exception, either homo- or heterozygotes for 136-V, as were four of the five sheep with clinical scrapie. The incidence of the VRQ allelic variant in the flock was unusually high compared to the Icelandic sheep population probably contributing to the high prevalence of both clinical and subclinical infection in the flock. Neither sheep with definite scrapie nor detectable subclinical infection, were of the resistant AHQ genotype, indicating that Icelandic AHQ sheep are not healthy carriers of scrapie infection.