Theoretical approaches predict that host quantitative resistance selects for pathogens with a high level of pathogenicity, leading to erosion of the resistance. This process of erosion has, however, rarely been experimentally demonstrated. To investigate the erosion of apple quantitative resistance to scab disease, we surveyed scab incidence over time in a network of three orchards planted with susceptible and quantitatively resistant apple genotypes. We sampled Venturia inaequalis isolates from two of these orchards at the beginning of the experiment and we tested their quantitative components of pathogenicity (i.e., global disease severity, lesion density, lesion size, latent period) under controlled conditions. The disease severity produced by the isolates on the quantitatively resistant apple genotypes differed between the sites. Our study showed that quantitative resistance may be subject to erosion and even complete breakdown, depending on the site. We observed this evolution over time for apple genotypes that combine two broad-spectrum scab resistance QTLs, F11 and F17, showing a significant synergic effect of this combination in favour of resistance (i.e., favourable epistatic effect). We showed that isolates sampled in the orchard where the resistance was inefficient presented a similar level of pathogenicity on both apple genotypes with quantitative resistance and susceptible genotypes. As a consequence, our results revealed a case where the use of quantitative resistance may result in the emergence of a generalist pathogen population that has extended its pathogenicity range by performing similarly on susceptible and resistant genotypes. This emphasizes the need to develop quantitative resistances conducive to trade-offs within the pathogen populations concerned. (C) 2014 Elsevier B.V. All rights reserved.