Very little is known about the biology and ecology of haploid Armillaria strains in nature. In this outdoor inoculation experiment, we assessed the virulence of six haploid Armillaria ostoyae strains along with their diploid parent towards 2-year-old seedlings and 4-year-old saplings of Norway spruce (Picea abies), and determined their ability to colonise freshly cut stumps. As inoculum source an Armillaria-colonised hazelnut (Corylus avellana) stem segment was inserted into the soil substrate. Re-isolations from mycelial fans at the root collar of infected trees or stumps were made. Surprisingly, not a single haploid re-isolate could be recovered. Microsatellite genotyping of 133 re-isolates suggests that the inoculated haploid strains were diploidised either by mating propagules (basidiospores or haploid mycelia) already present in the soil substrate or naturally disseminated in the course of the experiment from nearby forests. Consequently, no conclusion about the infectious ability of haploid Armillaria mycelia under natural conditions can be drawn. Nonetheless, the diploid half-sib families resulting from the diploidisation showed varying degrees of virulence, with a high correlation between the experiment with 2-year-old seedlings and 4-year-old saplings. Despite extensive genotyping of re-isolates, no evidence for somatic recombination between haploid mating propagules and diploidised mycelia was detected, suggesting that this is an uncommon phenomenon in A. ostoyae.