Among a population of Thecamoeba quadrilineata (Thecamoebidae) isolated from moss samples some trophozoites harboured intracellular aggregates of round-oval parasites which turned out to be spores with a real nucleus. These organisms were supposed to be fungal endoparasites beginning their development within the nucleus of the host invaded by young parasitic stages after the host amoeba had engulfed free spores from the environment. The complete developmental cycle was studied by electron microscopy, showing the intranuclear growth of freshly invaded young stages into large spore forming parasites differentiating into a great number of spores – all within the border of the host’s nuclear membrane. These spores were not released into the environment until the death and decay of the host amoeba, where they could be ingested as infective stages by hitherto not infected thecamoebae. Host range studies with various free-living amoebae (FLA) showed that T. striata and T. terricola were as permissive to infection as the original host T. quadrilineata. Sappinia was only susceptible to a certain extent and therefore not considered as possible natural host. Remarkably, this observation shows that both nuclei of the bi-nucleate amoeba became simultaneously infected. The present morphological description corresponds to early observations with fungal intranuclear parasites called Nucleophaga Dangeard, 1887. However, genetic and phylogenetic studies have to corroborate the supposed fungal nature.