The story began with the first bones. In 1887, Othniel Charles Marsh's first find was a pair of Triceratops horn cores, the bony inner part of the horns. Marsh first thought the horn bones belonged to a giant bison, and he named Triceratops in 1889. The skull he used to name it belongs to Yale's Peabody Museum of Natural History. Collecting followed. John Bell Hatcher found and dug up more than 30 Triceratops skulls for Yale in two years, and a Triceratops skull was collected in 1891 on a Yale expedition in the Lance Formation of Wyoming that he led. In 1905 the Smithsonian mounted the first Triceratops skeleton, put together from 10 specimens from various places in Niobrara County. Not every skull survived its trip: a Yale Triceratops skull sent to the Netherlands was apparently dropped on arrival and smashed into hundreds of pieces.

Triceratops is one of the most common dinosaurs found as fossils, and hundreds of skulls have been found in Montana and the Dakotas, but usually without the rest of the skeleton. That abundance invites a hard question: what does a difference between two skulls mean? By 1949, as many as 16 species of Triceratops had been named, because nearly every difference between skulls was treated as grounds for a new species. Scientists today count two, T. horridus and T. prorsus. Allowing for variation within species, including change with growth, cut 18 latest Cretaceous horned dinosaur kinds down to two, in the view of Scannella and Horner. The central problem is telling a species apart from a growth stage.

The tools for that problem are statistical, and they are open to criticism. Researchers coded Torosaurus and Triceratops skulls for traits that change with maturity and used clustering to arrange them into a growth series. Longrich and Field found that adult body size varies a lot in horned dinosaurs, so size cannot be used as a stand-in for maturity or to tell species apart. They suggested that bone remodeling may simply not be a reliable indicator of maturity. They also criticized one growth analysis because it plotted the squamosal's length-to-width ratio against its length, so length entered both terms and exaggerated the correlation. In 2009, Scannella and Horner reported that fossils called Torosaurus may be Triceratops that reached full adulthood before dying, while in 2012 Longrich and Field concluded that Torosaurus is a distinct genus of horned dinosaur. Fewer than a dozen adult Torosaurus specimens have been found, compared with many dozens of younger Triceratops.

Rock layers offer a different test, because they add time. A 2014 study placed more than 50 Triceratops skulls in the rock layers of Hell Creek, Montana, and found that the T. horridus form was low in the formation and the T. prorsus form in the upper third. The authors said the separation makes the idea that they are sexes or growth stages of one species untenable. Skulls from the middle of the formation show mixed traits, consistent with anagenesis, the transformation of one lineage over time. The authors also noted that variation in specimens from the middle zone may point to a branching event in the Triceratops lineage. That is the other idea: that Triceratops split from an earlier dinosaur into two lineages, called cladogenesis, which would mean both species lived at the same time. A 1992 study had suggested one lineage changing into the other but used too few samples.

Sample size matters just as much in the study of injuries. A 2009 study of museum skulls found far more lesions on the frill's squamosal bone in Triceratops than in Centrosaurus, a horned dinosaur with a big nose horn and small brow horns. It used more than 400 observations, and the most common lesion was periosteal reactive bone: 22 of the 26 lesions it counted. To increase sample size, the team treated isolated skull pieces from the Hell Creek and Lance Formations as Triceratops, since it is overwhelmingly the most common horned dinosaur there. They left out juvenile and subadult skulls, to control for possible changes in behavior as the animals grew up. The authors ruled out predator attacks as the main cause, because tyrannosaurs lived in the habitats of both animals. They did not claim a precise cause for any one injury, saying only that the overall pattern fits fighting among Triceratops. Farke's team said one idea worth further consideration is that Centrosaurus lost long brow horns or changed combat styles to reduce skull injury.

Teeth are a more physical kind of evidence. Dinosaurs evolved precise tooth-to-tooth chewing four times, in heterodontosaurid ornithopods, hadrosaurids, giant sauropods and ceratopsians. Most reptile teeth are used for grabbing and crushing and do not meet like mammal teeth, so reptiles cannot really chew. Coronal cementum in Triceratops teeth is only the second case of that tissue evolving separately in reptiles; hadrosaurs are the other. The vasodentine in Triceratops teeth was so porous that it wore unusually fast, which formed fullers, the grooves that cut friction on the chewing surface. Tests on 66-million-year-old teeth showed that their wear properties were still preserved, which let researchers build a 3D wear model.

The nose carries the newest open question. As of November 2026, University of Tokyo researchers have used CT scans of fossil Triceratops skulls, compared with the snouts of birds and crocodiles, to work out what filled the huge nasal space. As of November 2026, in Triceratops the skull blocks the jaw route that nerves and blood vessels use in most reptiles, so they reach the nostrils by the nasal branch. Respiratory turbinates are thin, scroll-shaped bones in the nasal cavity that let blood and air exchange heat more efficiently, and as of November 2026 researchers think Triceratops had them. As of November 2026, the Tokyo team said they are not 100% sure Triceratops had a respiratory turbinate, but some birds have an attachment ridge for one and horned dinosaurs have a similar ridge in the nose. As of November 2026, Triceratops was likely not fully warm-blooded, but nasal turbinates may still have helped it control temperature and keep moisture.

Words to know

• anagenesis: one lineage changing
• cladogenesis: a lineage splitting
• turbinates: scroll-shaped nose bones
• lesions: damaged spots
• vasodentine: porous tooth tissue