The "hunt for Praxiteles" involves identifying the most reliably documented statue types, described in the source texts, depicted on coins, and copied by later generations.

The Apollo Sauroktonos ("the Lizard-Slayer")

"Apollo the Lizard-Slayer" is without doubt the best-known statue type attributed to Praxiteles after the Aphrodite of Cnidus, thanks to the existence of numerous literary and numismatic sources (some 20 Roman replicas exist, in marble and bronze). Its distinctive subject-matter permits a confident identification of the work, and a sound attribution to Praxiteles, even though the statue's original meaning, function and location remain unknown. Pliny writes that Praxiteles executed a number of fine works in bronze: "He made an Apollo, at the age of puberty, who with an arrow in his hand is poised to strike a lizard climbing towards him: it is known as the Lizard-Slayer."

The group features a standing naked youth, and a lizard climbing a tree. Apollo places his weight on his right leg and throws out his right hip. His torso is turned slightly downward, to the right and his left arm is lifted toward the front. He leans his left arm against the tree, while the action of his right arm (holding the arrow) causes him to pivot slightly at the waist. His hair is shown in a highly sophisticated arrangement.

The figure is the subject of considerable expert debate: is this Apollo an adolescent (as described by Pliny) or a child (as described by the Roman poet Martial)? The existing copies fall into two main types: one in which the figure stands close to the tree-trunk, as in the version in the Vatican, and the other further away (as in the Louvre figure). Representations of the figure on coins would seem to confirm that the Louvre figure is the closest to Praxiteles's original. However, it is also argued that the Vatican figure (standing close to the tree-trunk) is technically less advanced than the figure in the Louvre, whose composition demanded a more sophisticated structural approach, in order to ensure its stability. In this context, the Vatican type might well be the older of the two, and closer to the original. Looked at together, the various versions of the Apollo Sauroktonos would seem to indicate that the copyists gradually conflated this statue type with that of the Praxitelean Eros, with numerous variations in the representation of the lizard, and no single predominant scheme.

The context surrounding the statue's creation is still more obscure: for whom was it made, and where was it located? The numismatic sources tell us little: coins from Apollonia in Mysia show Apollo minus the lizard altogether; Apollo is represented as a lizard-slayer on coins from Nicopolis, but the city was founded five centuries after the statue was first created, under the Roman Emperor Trajan. The iconographical type of Apollo as the "Lizard-Slayer" has no textual source, but this has not deterred numerous fanciful interpretations of the figure.

Praxiteles and other artists of his time often presented the gods as youthful, sexually ambiguous figures. This Apollo can be seen as the masculine pendant to Praxiteles's experimentations in the representation of nudity, paralleling the Aphrodite of Cnidus for the naked female form. In the wider context of Praxiteles's œuvre, the typology of the Apollo Sauroktonos would seem to lie mid-way between the Pouring Satyr of circa 370 BC (which remains faithful to the Polykleitian type) and the more supple, leaning figure of the Olympian Hermes, executed circa 340 BC.

Some commentators have rejected the figure's attribution to Praxiteles, identifying it instead as a pastiche of a Greek work, made for a Roman client in the 1st century BC by a certain Pasiteles. The creation of a sculpture of this type in the 4th century BC is not impossible however: votive reliefs show that Greek sculptors studied similar attitudes, incorporating trees into their compositions. The strong emphasis on the frontal viewpoint, the taste for sharply defined contours, and the piece's "closed," self-sufficient composition all support the plausibility of an attribution to Praxiteles.