Lambeosaurus (/ˌlæmbiəˈsɔːrəs/ LAM-bee-ə-SOR-əs) is a genus of hadrosaurid dinosaur that lived during the Late Cretaceous period of western North America. The first skull of Lambeosaurus found was used by palaeontologist Lawrence M. Lambe to justify the creation of the new genus Stephanosaurus, although it was not part of the latter's original material. Its incomplete nature led William A. Parks to name Lambeosaurus lambei for this skull in 1923 to honour Lambe. Multiple species of Lambeosaurus have been named since, including L. clavinitialis and L. magnicristatus in 1935, and L. laticaudus in 1981 which was later moved to its own genus Magnapaulia. It has also been identified that some species earlier identified as belonging to Tetragonosaurus and Corythosaurus are now considered juveniles of Lambeosaurus. It is the eponymous member of the subfamily Lambeosaurinae and tribe Lambeosaurini. Lambeosaurins, which also includes Corythosaurus and Hypacrosaurus from western North America, are understood to be some of the most specialized ornithopods.

Adult Lambeosaurus would have grown to around 7–7.7 m (23–25 ft) long and weighed 2.6–3.4 t (2.6–3.3 long tons; 2.9–3.7 short tons). It was able to move on two or four legs, with a deep tail, long limbs, and a highly distinct, hollow cranial crest. This crest, which can be used to separate the three recognized species of Lambeosaurus, projects well above the eye and slightly over the snout, and adults of some species possess a backwards spur. The function of the crest, which is also found in other lambeosaurines, is debated historically, but modern studies show that it could have been used as a resonating device for vocalisation, with a secondary function of sexual or species identification. The crest also allows for the identification of juveniles, which are otherwise nearly indistinguishable from those of Corythosaurus. It is through this identification that the growth of Lambeosaurus is well-known, with the crest developing late but expanding in height by an order of magnitude by the time individuals reached adulthood. Skin impressions are known and show that it had unornamented scales across the entire body.

The species of Lambeosaurus are only known from the middle Campanian of the Dinosaur Park Formation of Alberta. L. clavinitialis is known from a restricted range as the oldest species, overlapping with L. lambei which lived for around 0.3 million years before L. magnicristatus evolved later in the Campanian. This temporal separation suggests that L. clavinitialis, which was previously believed to be a female of either L. lambei or L. magnicristatus, is a separate species or at least earlier population. Lambeosaurus would have lived alongside the lambeosaurines Corythosaurus and Parasaurolophus, and also the saurolophine Prosaurolophus. Dental wear suggests that Lambeosaurus would have avoided competition with Prosaurolophus by occupying different feeding niches, preferring more closed habitats and browsing lower to the ground with a more generalised diet. The habitat Lambeosaurus lived in was a coastal plain where meandering rivers separated regions of dense vegetation, covered in a diversity of conifers, ferns and other shrubs, and occupied by diverse invertebrates, fish, mammals and reptiles, especially other megaherbivorous dinosaurs.

In the 1880s and 1890s, expeditions of the Geological Survey of Canada into Alberta discovered that the rocks along the Red Deer River bore dinosaur fossils of scientific importance. These deposits were identified as belonging to either the Edmonton or Belly River Series, of the middle to end Cretaceous. Canadian palaeontologist Lawrence M. Lambe undertook three expeditions in 1897, 1898 and 1901 to an extensive series of badlands between Berry Creek and Deadlodge canyon. The fossils discovered were fragile and therefore difficult to excavate, but belonged to many species, including three new species of the hadrosaur Trachodon that Lambe named Trachodon selwyni, Trachodon marginatus, and Trachodon altidens in 1902. T. marginatus was named for a partial skeleton as well as isolated jaw and limb bones, and T. altidens was named for a partial maxilla with many teeth. In the same publication, American palaeontologist Henry Fairfield Osborn summarized the fauna of the mid-Cretaceous across North America, and provided the possible new subgenus name Didanodon for T. altidens.

A 1913 expedition of the Geological Survey to the same location resulted in the discovery of multiple hadrosaur skeletons by American palaeontologist Charles Hazelius Sternberg. Among these specimens was a skull, skeletons, and skin impressions, which Lambe regarded as additional material of T. marginatus in 1914. One of these specimens (CMN 351), found by Sternberg 5.6 km (3.5 mi) southeast of the mouth of Berry Creek, was described by Lambe more thoroughly later in 1914. Based on this specimen, Lambe concluded that T. marginatus belonged in a new genus, which he named Stephanosaurus. To Stephanosaurus he assigned the original material of T. marginatus, as well as CMN 351 and another specimen found by Sternberg in 1913. Lambe also attempted to replace the holotype of S. marginatus (CMN 419), the partial skeleton he described in 1902, with the isolated jaw bones (CMN 361 and 362) he described in the same year, which is not permitted by the rules of zoological nomenclature. However, American palaeontologist Barnum Brown argued later that year that the skull and jaws cannot be confidently assigned to Stephanosaurus because the type specimen does not include any skull material to compare with, and noted similarities with the skull of his new genus Corythosaurus. In 1920, Lambe assigned another, even more complete skull (CMN 2869) to Stephanosaurus that showed differences to Corythosaurus. This skull was found by Charles Mortram Sternberg (son of C.H. Sternberg) in 1917 from around 6.4 km (4 mi) southeast of the mouth of Little Sandhill Creek. However, Canadian palaeontologist William A. Parks noted that Brown's logic still applied, and designated the new genus and species Lambeosaurus lambei for the complete skulls CMN 351 and CMN 2869 in 1923, as they could not justifiably be assigned to Stephanosaurus. Parks chose the name Lambeosaurus lambei to give Lambe, who had died in 1919, as much credit as possible for the initial identification of new hadrosaur.

In 1924, American palaeontologist Charles W. Gilmore described the type material of Lambeosaurus in more detail. He found that the type material of Stephanosaurus was likely not the same taxon as the skulls and may have belonged to Kritosaurus, and that the two jaw specimens Lambe tried to designate as types were also possibly Kritosaurus. Gilmore selected the better preserved skull that Lambe described in 1920 (CMN 2869) as the type specimen of Lambeosaurus lambei, as no type specimen had been designated by Parks. Additional specimens that Gilmore assigned to Lambeosaurus include CMN 351 and CMN 8503, the latter of which includes a partial skull and articulated skeleton also found by C.M. Sternberg in 1917, 5.6 km (3.5 mi) west of the mouth of Little Sandhill Creek. Other early material of Lambeosaurus includes one of the specimens collected by C.H. Sternberg in 1913 that is now at the University of British Columbia, where it was put on display in 1950; a skull (FMNH UC 1479) collected in 1926 by Levi Sternberg (son of C.H. Sternberg); and a skeleton missing the skull (FMNH PR 380) collected in 1922 by an expedition of the Field Museum of Natural History, which was originally identified as Prosaurolophus before being prepared and mounted as a Lambeosaurus being scavenged by Daspletosaurus. It was earlier believed that the skull and skeleton could have belonged to the same individual, but it is now known that they were found 9.7 km (6 mi) apart.

The American Museum of Natural History also excavated in the Red Deer River region, primarily through the work of Brown from 1909 to 1914. One particularly small individual discovered by Brown, was assigned to as "Procheneosaurus" by American palaeontologist William Diller Matthew in 1920, and then again in 1923 by Matthew and Brown where the specimen in question was identified as AMNH 5340. Matthew himself did not believe that the description was adequate to name a new taxon, writing to C.M. Sternberg in 1921 that "Procheneosaurus" should be considered a nomen nudum, an informal name not to be used in taxonomy. In 1931, Parks described more small, crested hadrosaurs found by expeditions of the University of Toronto into the Red Deer badlands. The first (ROM 3577) was found around 3.2 km (2 mi) southeast of Little Sandhill Creek by L. Sternberg. Parks described this specimen, which includes a skull and part of the vertebral column, as the new taxon Tetragonosaurus praeceps. The second specimen, ROM 3578, was found in 1927 by L. Sternberg around 2.4 km (1.5 mi) downriver of Little Sandhill Creek. This specimen, which comprises only the skull, was named Tetragonosaurus erectofrons by Parks. Together, both species were considered to be close to Cheneosaurus and had similar low-domed crests.

In 1935, Charles M. Sternberg reassessed the taxonomy of crested hadrosaurs after completing the preparation of the 18 skulls and skeletons at the Canadian Museum of Nature. This work resulted in the naming of two new species of Lambeosaurus, L. clavinitialis and L. magnicristatum, and one new species of Tetragonosaurus, T. cranibrevis. The latter was based on a partial skull (CMN 8633) that Sternberg found in 1928 ca. 3.62 km (2.25 mi) south of the mouth of Berry Creek. L. clavinitialis was named for a skull and skeleton (CMN 8703) that Sternberg found nearby, also in 1928, ca. 4.0 km (2.5 mi) south of the mouth of Berry Creek. L. magnicristatum was named for a mostly complete skull and skeleton (CMN 8705) Sternberg found in 1919, about 4.8 km (3 mi) southwest of the mouth of Little Sandhill Creek. Sternberg also reassigned the specimen CMN 8503 to Corythosaurus; this specimen was previously assigned to L. lambei by Gilmore. Also in 1935, Parks named the new species Corythosaurus frontalis for a specimen from the same area (ROM 869).