In this paper some clinical fragments are presented to illustrate the delicate relationship between initial diagnostic assessment and the phenomenology that becomes visible through the gradual unfolding of an analytic cure. Examples of gerontophilia, apparently extreme narcissism, fetishism, and frotteurism illustrate that a theoretical diagnosis of perversion is far from straightforward. A cure, which is carried out thoroughly and which is faithful to its own dynamics, requires careful handling of hypothetical psychodiagnostic categories, even psychoanalytically inspired ones. Perverse onsets or perverse traits are not necessarily indications of a perverse structure.