In book VI of his Parmenides commentary, Proclus surveys the previous Neoplatonic interpretations of the “hypotheses” of Plato’s Parmenides, starting with Amelius, a disciple of Plotinus. This paper analyses Amelius’ exegesis of the hypotheses. After presenting the extant evidence about his Parmenides commentary, I explore three puzzling aspects of his interpretation: 1) unlike most Neoplatonists, Amelius distinguished eight hypotheses instead of nine. I argue that he considered the last section (Prm. 165e2-166c2) to be devoid of any positive content and merged (...) it with the previous one – contrary to the modern advocates of an eight-fold divisio textus. 2) Amelius assumed that each hypothesis deals with one layer of reality, but the precise meaning of some identifications is elusive: what are the three kinds of matter matched to hypotheses 5-7? I show that these three material layers probably correspond to three entities whose existence is posited on the basis of the Timaeus: the chōra/prime matter, the level of the element-traces, and the ordered universe. 3) In at least one point, Amelius’ exegesis disrupts the descending hierarchical ordering of the subject-matters of the hypotheses (i.e. from the highest to the lowest-ranked entity), which was an undisputed governing principle of the Neoplatonic interpretations. This irregularity depends, as I argue, on a further interpretative rule: the negative quality of a hypothesis (except for the first one) is taken to indicate some sort of debasement of the corresponding entity. Since, in the second part of the Parmenides, the negative hypotheses alternate with the positive ones, this further principle clashed with the linear downward arrangement implied by the hierarchical principle. In the conclusion of the paper, I sketch how the insights gained by this analysis can shed some light on the subsequent development of the ancient Parmenides exegesis. (shrink)