Elopteryx nopcsai is a Maastrichtian paravian from Romania known principally from incomplete femora. Since 1913 it has been interpreted as a cormorant-like bird, an indeterminate theropod, a possible alvarezsaurid and a basal avialan. A femoral fragment described in 2024 strengthened the terrestrial bird interpretation and revived a proposed connection with Balaur, but the available bones still do not resolve its identity.
Quick facts
| Scientific name | Elopteryx nopcsai Andrews, 1913 |
|---|---|
| Group | Theropoda, Maniraptora, Paraves; exact placement disputed |
| Age | Maastrichtian, near the end of the Cretaceous |
| Range | Transylvania, Romania |
| Holotype | BMNH A PAL.1234, proximal fragment of a left femur |
| Other material | Femoral fragments BMNH A PAL.1235 and BBU-PSM V1009 are referred or compared |
| Size | No reliable whole-body length or mass |
| Diet | Unknown |
How secure is each part of the reconstruction?
BMNH A PAL.1234 is an incomplete upper end of the left femur collected near Sânpetru.
BBU-PSM V1009 resembles the holotype and has thick cortical bone, but was found separately.
UBB V649 has a broad fused sacral region and resembles Balaur, but no direct connection to the type exists.
The head, arms, feathers, feet and tail are unknown, and even the full proportions of the leg cannot be measured.
Nopcsa’s fossils and the 1913 name
Baron Franz Nopcsa collected fossils around Sânpetru in the late nineteenth and early twentieth centuries. His material included two proximal femoral fragments and an incomplete lower leg bone. Exact coordinates and associations were not recorded, so similar size and surface texture cannot demonstrate that all pieces came from one animal or species.
Charles William Andrews named Elopteryx nopcsai in 1913. He selected BMNH A PAL.1234 as the holotype, referred BMNH A PAL.1235 to the species and assigned the lower leg fragment provisionally. The generic name combines Greek roots for marsh and wing; the species honours Nopcsa. Andrews compared the animal with birds then grouped near cormorants, while recognising its unusually terrestrial build.
Why the material became mixed
Romanian distal leg bones moved repeatedly among Elopteryx, the supposed Cretaceous owls Bradycneme and Heptasteornis, ornithomimosaurs and alvarezsaurs. A taxon based on one femoral fragment cannot automatically absorb every similar leg bone from the same formation.
A 2005 review described the separate distal femur FGGUB R.351 from La Scoabă and tentatively compared it with alvarezsaurids. Its exact position in the section was also unknown. These assignment problems are not minor bookkeeping: combining bones from different animals can manufacture a body plan that no individual possessed.
What the 2024 femur contributes
BBU-PSM V1009 is the proximal part of a right femur recovered from a concretion at Nălaț-Vad in lower Maastrichtian beds of the Sânpetru Formation. Its upper end resembles the Elopteryx holotype. A cross-section shows thick cortical bone, which the authors associated with a robust ground-dwelling animal and possible secondary loss of flight.
That conclusion has two dependencies. The bone must first belong to Elopteryx, and Elopteryx must occupy the bird line before the word “secondary” applies. No wing, shoulder girdle or feather impression provides a direct flight test.
The pelvis and Balaur hypothesis
The partial pelvis UBB V649, found at Nălaț-Vad in 2003, preserves at least nine fused sacral vertebrae, a broad synsacrum and a developed antitrochanter. Researchers first compared it with Gargantuavis, while also noting similarity to Balaur and size compatibility with Elopteryx femora.
Neither the pelvis nor the 2024 femur was articulated with the 1913 holotype. More importantly, the Balaur type skeleton lacks femora, while the Elopteryx type lacks Balaur’s diagnostic hands and feet. The two name-bearing specimens therefore have no overlapping bone. Synonymy is possible, and Elopteryx would have priority if proved, but compatible proportions cannot prove it.
Age, habitat and unknowable details
The principal fossils come from Maastrichtian continental deposits of Transylvania. Nălaț-Vad records a river floodplain with fine overbank sediments, caliche horizons and evidence of seasonal conditions. Amphibians, turtles, crocodylomorphs, dinosaurs, pterosaurs, birds and mammals lived in the wider region, but association in one basin does not establish Elopteryx’s precise habitat or interactions.
Femoral fragments show parts of the hip joint and muscle attachments but cannot produce a secure body length or mass. The skull, jaws, teeth, beak, stomach contents and coprolites are absent, so carnivory, omnivory and herbivory are all untested. Plumage is plausible for a paravian, whereas feather shape, wings, tail, colour, speed and social behaviour are reconstruction.
Evidence and reconstruction
| Level | What belongs here |
|---|---|
| Direct | The proximal left femur BMNH A PAL.1234 and its preserved anatomical surfaces |
| Supported inference | A robust hind limb and probable membership in Paraves |
| Disputed | Avialan position, referral of other bones, flightlessness and synonymy with Balaur |
| Unknown | Whole-body size, head, arms, feet, diet, feathers, colour and behaviour |
Frequently asked questions
Was Elopteryx a bird?
An avialan identity is a serious hypothesis, but the fragmentary bones have produced several positions in different studies. Its precise place within Paraves remains unresolved.
Which fossil definitely belongs to Elopteryx?
The name is securely attached to BMNH A PAL.1234, the proximal end of a left femur. Other femoral and pelvic material has been compared or referred but was not found with the holotype.
Was Elopteryx flightless?
A thick-walled referred femur supports a robust terrestrial interpretation, but no wing or feathers are known. Flightlessness also depends on whether that bone belongs to Elopteryx and whether the genus is an avialan.
Are Elopteryx and Balaur the same animal?
Possibly, but not demonstrably. Their holotypes preserve no matching bone, so proposed similarities in separately found material cannot yet establish synonymy.

