Attempts were made to reproduce avian vacuolar myelinopathy (AVM) in a number of test animals in order to determine the source of the causative agent for birds and to find a suitable animal model for future studies. Submerged vegetation, plankton, invertebrates, forage fish, and sediments were collected from three lakes with ongoing outbreaks of AVM and fed to American coots (Fulica americana), mallard ducks and ducklings (Anas platyrhynchos), quail (Coturnix japonica), and laboratory mice either via gavage or ad libitum. Tissues from AVM-affected coots with brain lesions were fed to ducklings, kestrels (Falco sparverius), and American crows (Corvus brachyrhynchos). Two mallards that ingested one sample of Hydrilla verticillata along with any biotic or abiotic material associated with its external surface developed brain lesions consistent with AVM, although neither of the ducks had clinical signs of disease. Ingestion of numerous other samples of Hydrilla from the AVM affected lakes and a lake with no prior history of AVM, other materials (sediments, algae, fish, invertebrates, and water from affected lakes), or tissues from AVM-affected birds did not produce either clinical signs or brain lesions in any of the other test animals in our studies. These results suggest that waterbirds are most likely exposed to the causative agent of AVM while feeding on aquatic vegetation, but we do not believe the vegetation itself is the agent. We hypothesize that the causative agent of AVM might either be accumulated by aquatic vegetation, such as Hydrilla, or associated with biotic or abiotic material on its external surfaces. In support of that hypothesis, two coots that ingested Hydrilla sampled from a lake with an ongoing AVM outbreak in wild birds developed neurologic signs within 9 days (ataxia, limb weakness, and incoordination), and one of two coots that ingested Hydrilla collected from the same site 13 days later became sick and died within 38 days. None of these three sick coots had definitive brain lesions consistent with AVM by light microscopy, but they had no gross or histologic lesions in other tissues. It is unclear if these birds died of AVM. Perhaps they did not ingest a dose sufficient to produce brain lesions or the lesions were ultrastructural. Alternatively, it is possible that a separate neurotoxic agent is responsible for the morbidity and mortality observed in these coots.