It was closely related to Stegosaurus and was similar to it in having two rows of, possibly alternating, plates on its back and four spikes on its tail end.

As he had difficulty securing the specimen due to the hard rock matrix, he sought help from Ronald G. Mjos and Jeff Parker of Western Paleontological Laboratories, Inc.

The type species Hesperosaurus mjosi was named and described in 2001 by Kenneth Carpenter, Clifford Miles, and Karen Cloward.

The specific name honours Mjos who, apart from his involvement in the process of collecting and preparing the holotype, also had a cast of it made, exhibited with the inventory number DMNH 29431 in the Denver Museum of Natural History.

[1] The holotype, HMNH 001 (later HMNS 14), was found in the Windy Hill Member, stratigraphic zone 1 of the lower Morrison Formation,[2] dating from the early Kimmeridgian, about 156 million years old.

In 2015, the HMNS permanently closed and the holotype was transferred to the Fukui Prefectural Dinosaur Museum, where it was renumbered FPDM-V9674[3] From 1995 onward at the Howe-Stephens Quarry in Big Horn County, Wyoming, named after the historic location of the Howe Ranch, once explored by Barnum Brown, and the new owner Press Stephens, Swiss palaeontologist Hans Jacob Siber excavated stegosaur specimens.

A third specimen was found in 2002: SMA L02, dubbed "Lilly" after the sisters Nicola and Rabea Lillich assisting the excavations as volunteers.

However, Susannah Maidment and colleagues in 2008 published a more extensive phylogenetic study in which it was recovered as a derived form, closely related to Stegosaurus and Wuerhosaurus.

Due to his conclusion that Hesperosaurus were rather basal, in it many comparisons were made with the basalmost known stegosaurian Huayangosaurus,[1] that lost their relevance once it became clear that the phylogenetic position was in fact quite derived.

Originally, Carpenter reconstructed the disarticulated skull elements into a very convex head, modelling it on the shape of Huayangosaurus.

[1] The discrepancies in the vertebral count are caused by applying different criteria to the problem whether (and which) cervicodorsal vertebrae should be considered part of the neck or the back.

[1] Peter Malcolm Galton in 2007 established some differences: there are rough vertical ridges present on the upper part of the crown, one per denticle; the fine grooves on the tooth surface are weakly developed.

[1] In 2012, an histological study concluded that these osteoderms, skin ossifications, of Hesperosaurus are essentially identical in structure to those of Stegosaurus.

CAT-scans showed that the plates have thin but dense outer walls, filled with thick spongy bone.

Additionally on some areas a black layer is present, possibly consisting of organic remains or bacterial mats.

A part of the lower trunk flank shows rows of small hexagonal, non-overlapping, convex scales, two to seven millimetres in diameter.

Apart from the scales, an impression of the lower side of a back plate has been found, covering about two hundred square centimetres.

Also the display function would have been reinforced, because the sheath would have increased the visible surface and such horn structures are often brightly coloured.

[5] In 2001 Carpenter performed a cladistic analysis showing that Hesperosaurus was rather basal and related to Dacentrurus:[1] Huayangosaurus ?Chungkingosaurus Chialingosaurus Wuerhosaurus Dacentrurus Hesperosaurus Tuojiangosaurus Kentrosaurus Lexovisaurus Stegosaurus stenops Stegosaurus ungulatus Carpenter was aware that his analysis was limited in scope.

[1] More extensive phylogenetic studies by Maidment recovered Hesperosaurus as a very derived stegosaurid, and the sister species of Wuerhosaurus.

The position of Hesperosaurus in the stegosaurid evolutionary tree according to a study from 2009 is shown by this cladogram:[15] Kentrosaurus Loricatosaurus Dacentrurus Miragaia Stegosaurus Wuerhosaurus (=Stegosaurus homheni) Hesperosaurus (=Stegosaurus mjosi) In a 2019 re-evaluation of dacentrurine stegosaurids, Costa and Mateus suggested that, based on their revised diagnosis for the clade Dacentrurinae, Hesperosaurus appears to heave been closely related to Dacentrurus after all, though they refrained from formally reassigning it to that group pending the completion of an expanded phylogenetic analysis.

[16] In 2015, a study by Evan Thomas Saitta based on the finds in the JRDI 5ES Quarry concluded that Hesperosaurus showed sexual dimorphism.

Though the back plates of the various individuals were not articulated, Saitta managed to order them into cervical, dorsal and caudal series for each type.