The rationale for dividing the clinical spectrum of DSM-III-R male heterosexual gender identity disorder into three types was examined. The DSM-III-R category of fetishism for female attire, was included in the analysis.There were 266 male participants divided into three groups: 172 fetishists for female attire or gender identity patients, 52 androphiles, and 42 gynephiles. A 16 item questionnaire was used to examine the groups. A three factor scale ("Fetishism", "Gender Dysphoria" and "Androphilia") was derived from the questionnaire. Only the Gender Dysphoria Factor Scale successfully differentiated between all four conditions, supporting the notion that the three types of gender identity disorder represent a continuum of degree of severity of gender dysphoria. Defining two of the three types of gender identity disorder in terms of the patients' self-reports on fetishism, as DSM-III-R does, is therefore unnecessary.adolescence or adulthood, non-transsexual type (GIDAANT); and transsexualism. These types were formulated on the basis of broad clinical experience, but until now the rationale for this division had not been formally tested. The present study conducted such a test and included, for reasons which will become evident later on, fetishism for female attire, also as defined by DSM-III-R A DSM-III-R essential characteristic of &dquo;transvestic fetishism&dquo; is &dquo;recurrent, intense, sexual urges and sexually arousing fantasies of at least six months duration, involving cross-dressing&dquo; (p.288). These sexually arousing fantasies clearly indicate the presence of fetishism. Conversely, a feature of GIDAANT is that &dquo;cross-dressing is not for the purpose of sexual excitement&dquo; (p. 76). Further, GIDAANT is without the presence of the diagnostic criteria for transsexualism as there is &dquo;no persistent preoccupation ... with getting rid of one's primary and secondary sex characteristics and acquiring the sex characteristics of the other sex&dquo; (p. 76).The differential diagnosis between fetishistic transvestism and GIDAANT has obviously been based on the self-reports of gender identity disorder patients as to the degree of their fetishism. However, self-reports of fetishism in patients of this sort should not be considered sufficiently reliable to be used in defining the types of this disorder. Playing down or denying fetishism is typical of a majority of these patients (Blanchard, 1988; Blanchard, Clemenssen, & Steiner, 1985; Blanchard, Racansky, & Steiner, 1986); they apparently want their transvestism to be viewed as resulting from a genuine psychological femaleness. However, this is not to say that these patients' self-reports of fetishism are of no use and should therefore be ignored.Self-reports about discontentment with gender appear to better serve the purpose of defining the types of heterosexual gender identity disorder. This is because: (a) This kind of self-report information is more directly connected to the patient's state than self-reports on fetishism. The more dysphoric t...