Mycelial fan formation was studied in five Armillaria cepistipes, ten A. borealis and ten diploid and six haploid A. ostoyae strains on excised stem segments of Picea abies. Stem segments were either non-autoclaved or autoclaved, representing dying and dead wood, respectively. To confirm the identity of mycelial fans on non-autoclaved stem segments, re-isolations were made and isolates characterized with microsatellite markers. Mycelial fan formation on autoclaved stem segments was fast and reliable for most of the tested Armillaria strains. On non-autoclaved stem segments, mycelial fan formation was slower, more erratic and less predictable. Mycelial fan formation was fastest in A. cepistipes closely followed by A. borealis and was slowest in A. ostoyae. For two A. cepistipes and four A. ostoyae strains (all diploid), growth rates of mycelial fans were estimated in a time course experiment. They ranged between 5.1 and 8.7 mm/day for autoclaved and between 1.4 and 4.7 mm/day for non-autoclaved stem segments. The haploid A. ostoyae strains also formed mycelial fans on autoclaved stem segments, but typically slower and less reliably than the diploid strains. Whether haploid strains are able to produce mycelial fans on non-autoclaved stem segments remains unknown because of accidental diploidization of the original haploid strains which was likely caused by basidiospores introduced into the study system on the non-autoclaved stems. Overall, the method developed in this study maybe useful for further investigations into the genetic, physiological and biochemical nature of mycelial fan formation in the genus Armillaria.