To assemble a laboratory population of Blumeria graminis f. sp. hordei for a competition experiment, controlled crosses were performed among 30 parent isolates, which were characterized by their pathotype (i.e., the phenotype observed on a differential set containing lines including host resistance alleles) and their response to the fungicides ethirimol and triadimenol. Variability in the response to the chemical inducer of host-resistance acibenzolar-S-methyl (BTH: benzo[1,2,3]thiadiazole-7carbothioic acid-S-methyl ester) was introduced by using isolates collected in fields repeatedly treated with this chemical. Based on their pathotype and response to ethirimol, 137 isolates recovered from the crosses were chosen to assemble a laboratory population. This protocol produced variability in the laboratory population for traits chosen only in the parents (triadimenol) or both in the parents and the progenies (ethirimol). It was therefore postulated that the variability in the response to BTH, if present in the parents, was also present in the laboratory population. No association between these traits was observed. The effect of BTH on the evolution of the laboratory population was compared with the effect of the fungicide ethirimol. The laboratory population was exposed to selection pressures and its evolution was followed over 10 generations. Ethirimol treatments always induced a decrease in sensitivity, whereas no consistent trend was observed for sensitivity to triadimenol (not selected). This result indicates that the application of ethirimol induced a selection pressure. For BTH, (toxicological) sensitivity tests have not detected any consistent evolution, but pathotype diversity indicated that in cases when BTH was applied with ethirimol, BTH induced a further selection pressure in addition to that of ethirimol.