Dracorex, an herbivore, had a skull with spiky horns, bumps, and a long muzzle. The species also sports well-developed supratemporal fenestrae and a heavily armored flat skull—lacking the characteristic pachycephalosaurid dome. Coupled with these two features is the excessive number of osteoderms in the form of irregular osteodermal crust: a number of nodes, larger hornlets, and spikes. Disregarding this, Dracorex is physically comparable to Stygimoloch.
In the Pachycephalosauridae, the Asian taxa include a number of (somewhat) flat-headed pachycephalosaurs (Homalocephale calathocercos, Goyocephale lattimorei, and Wannanosaurus yansiensis). However, prior to the discovery of Dracorex, the only semi-flat-headed pachycephalosaur from North America was Stegoceras validum (inclusive of Ornatotholus browni). Even then, the semi-flat-headed trait was only present in juveniles of the species.
Aside from having a flat, nodal skull, the most prominent feature of Dracorex is the pair of huge and unrestricted superior temporal openings. The supratemporal fenestrae are much larger front-to-back and side-to-side than in Homalocephale, and larger than in Goyocephale. Only a fragmentary Wannanosaurus skull shows fenestrae as large as those of Dracorex. This fenestral architecture has been seen in ancient archosaurs, but not in other pachycephalosaurs.
Consequently, if unreduced superior fenestrae are morphologically primitive, then Dracorex is more primitive in the temporal region than any other known pachycephalosaur. However, Sullivan (2003, 2006) demonstrated that the oldest known pachycephalosaurs were, in fact, fully domed, and that the flat-headed morphology appeared later in the fossil record. This suggests that doming may be primitive for pachycephalosaurs and that a reversion to the non-domed, flat-headed state is a secondary (derived) character reversion, coupled with the re-opening of the supratemporal fenestrae. Indeed, while Stegoceras has been considered to be transitional between domed and flat-headed taxa, it may indicate the beginning of a character reversion to suppression of doming and opening of the supratemporal fenestrae in some taxa.
The excavated specimen was most likely a young adult. However, based on the beginning of ossification of the mid-cervical arch with the centrum, it was near maturity. The animal was approximately 10 feet (3 m) long.
Dracorex may actually be an individual of the closely related Stygimoloch and/or Pachycephalosaurus in which the dome and horns are not well-developed, either because the animal was a juvenile or a female. This consideration was supported at the 2007 annual meeting of the Society of Vertebrate Paleontology. Jack Horner of Montana State University presented evidence, from analysis of the skull of the single existing Dracorex specimen, that this dinosaur may well be a juvenile form of Stygimoloch. In addition, he presented data that indicates that both Stygimoloch and Dracorex may be juvenile forms of Pachycephalosaurus. Horner and M.B. Goodwin published their findings in 2009, showing that the spike/node and skull dome bones of all three 'species' exhibit extreme plasticity, and that both Dracorex and Stygimoloch are known only from juvenile specimens, while Pachycephalosaurus is known only from adult specimens. These observations, in addition to the fact that all three forms lived in the same time and place, lead them to conclude that Dracorex and Stygimoloch may have simply been juvenile Pachycephalosaurs that had lost their spikes and grew domes as they aged. The researchers were unable to destructively sample the Dracorex skull, and had to use a cast of the skull for descriptive purposes. A 2010 study by Nick Longrich and colleagues also supported the hypothesis that all flat-skulled pachycephalosaurs were juveniles, suggesting that flat-skulled forms, like Goyocephale and Homalocephale, represent juveniles of dome-skulled adults.