From 12 cultivated and mostly vegetatively propagatedAllium species and varieties tested for carlavirus infections, 94 virus isolates were obtained which varied greatly on indicator hosts.Chenopodium amaranticolor, C. quinoa, Celosia argentea var.plumosa ‘Geisha’,Nicotiana hesperis accession 67A andN. occidentalis accession P1 proved valuable for detection, isolation and propagation of part of the isolates. The latter three species are new experimental hosts for carlaviruses ofAllium species. Other isolates could only be transmitted toAllium species such as crow garlic (A. vineale) leek (A. ampeloprasum var.porrum) and onion (A. cepa var.cepa). The isolates were grouped into three viruses by differential hosts and host reactions and their reaction with four antisera. Shallot latent virus (SLV) was found in ever-ready onion (A. cepa var.perutile), grey shallot (unidentifiedAllium species), multiplier onion (A. cepa var.aggregatum), pearl onion (A. ampeloprasum var.sectivum), rakkyo (A. chinense), shallot (A. cepa var.ascalonicum), and Welsh onion (A. fistulosum). Virus isolates from garlic and Asian shallot, fully reacting with antiserum to SLV but differing in host reactions from the SLV type-isolate, are now described as garlic strain (SLV-G) and Asian shallot strain of the virus, respectively. The ‘garlic latent virus’ from garlic described in Japan is now considered identical with SLV-G. A carlavirus almost universal in garlic, and also found in great-headed garlic (A. ampeloprasum var.holmense), in an unidentifiedAllium species, and occasionally in leek, did not react with the antisera to SLV and the Japanese ‘garlic latent virus’, and is now described as the new garlic common latent virus (GCLV). It appeared identical to a virus erroneously identified in Germany as garlic latent virus. The new Sint-Jan's-onion latent virus (SjoLV) from Utrechtse Sint-Jan's onion (unidentifiedAllium species) from the Netherlands and similar crops originating from other countries, did not induce reactions in test plants and could only be detected by electron microscope decoration tests. It reacted equally wellwith the antisera to SLV and GCLV. It was also present together with SLV in ever-ready onion, pearl onion, rakkyo, shallot, and Welsh onion. ‘Garlic latent virus’ reported in Japan from hosts other than garlic should be regarded as SLV, SjoLV, or a mixture of these viruses. The carlaviruses were not detected in wild plants of ramsons (A. ursinum), and of the predominantly vegetatively propagated crow garlic (A. vineale), field garlic (A. oleraceum), and sand leek (A. scorodoprasum), collected in the Netherlands. Severe reactions in the indicator hosts incidentally revealed soil-borne viruses in shallot (the nepovirusesArabis mosaic virus (ArMV) and tomato black ring virus) and crow garlic (ArMV and the tobravirus tobacco rattle virus). Tobacco necrosis virus (necrovirus) was detected in roots of shallot.