Three designs of field choice experiments were used to demonstrate complete reproductive isolation by sexually deceived thynnine pollinators (Neozeleboria Rohwer spp.) in eight species of Chiloglottis R.Br. Four species, Chiloglottis diphylla R.Br., C. formicifera Fitzg., C. pluricallata D.L.Jones and C. valida D.L.Jones, attracted only one wasp species, but the other four, C. platyptera D.L.Jones, C. seminuda D.L.Jones, C. trilabra Fitzg. and C. reflexa Labill. Druce exhibited multiple species attraction. Wasps visiting orchids were classified as major responders if they exhibited behaviour which could potentially result in pollination, by contrast to minor responders which did not. Major responders occurring sympatrically with the orchid were termed confirmed, potential or putative pollinators on the basis of observation of pollinia removal or deposition, pseudocopulation with the labellum, or arrival at bait flowers with pollinia from local orchid populations, respectively. Each Chiloglottis species had a single Neozeleboria species as confirmed or potential pollinator. The data confirmed that field pollinator choice tests can be used to distinguish cryptic sexually deceptive orchid species, and that specific pollinators may be used reliably as taxonomic characters in Chiloglottis.