Unbeknownst to the public and many scholars, the vast majority of ancient sculptural works - which are displayed in museums and cited in many textbooks as original and canonical pieces - are actually extensively restored eighteenth-century variations. Although during that time period extensive, and often creative, restoration was a well accepted practice, this does call into question the authenticity of these pieces and their usefulness as paradigms of ancient art.This is especially true for one of the most iconic and well known of ancient sculptural works - the Discobolus. The original bronze statue of a young nude discus thrower was created in the mid-fifth century BC by the Greek artist Myron. The original sculpture no longer exists. The extant works are all restorations of Roman marble copies with linkage to an extremely prolific eighteenth-century Italian sculptor and restorer, Bartolomeo Cavaceppi.This dissertation explores the work and practice of Cavaceppi during the early period of Neoclassicism, with special emphasis on the restoration of five different variants of the Discobolus. It begins with an examination of the original Greek statue and why so many Roman copies were made. The main focus, however, is on the variations that were introduced when the Roman pieces were restored between the sixteenth and eighteenth centuries, most notably those connected to the workshop of Cavaceppi. Using the Discobolus as a case study, my intention is to clarify the nature of these restored ancient pieces.