Gender assignment at birth in cases of ambiguous genitalia (AG) has long been based on anatomical appearance, adequacy of the phallus and/or response to testosterone. Recent reports have questioned the validity of this approach. We have encountered three patients born with AG, all with 46XY karyotype, whose gender assignments were made shortly after birth but later questioned or altered by the patients themselves. CASE 1: This patient, exposed to dilantin in utero, was born with a small phallus (considered to be an enlarged clitoris) and bilaterally descended testes in the labioscrotal fold. Female gender was assigned. No specific hormonal diagnosis was made; however, androgen insensitivity and 5α-reductase deficiency were ruled out. Gonadectomy was performed at 1 mo. of age. Starting at age 4 yr., the patient described herself as not feeling “like a girl”. She was described as “tomboyish” . At 12 y.o. she was started on estrogen but later discontinued the therapy. At age 17 yr. she declared herself male, changed her name, had mammoplasty and began testosterone therapy. CASE 2: An infant, born with AG and non-palpable gonads was assigned male gender at birth and raised as a boy. No specific diagnosis was made. Throughout childhood, he did not accept being told he was “a normal boy” . Hypospadias repair and gonadectomy were done at 5 y.o. At age 11 yr. testosterone was started but the patient did not accept the treatment and has been non-compliant since. Testicular prostheses were placed at 13 y.o. and removed at the patient's request at 18 y.o. Now, at age 27 yr., the patient identifies himself as “intersex” . CASE 3: This patient, born with AG and non-palpable gonads, was assigned female gender at birth because of the presence of a uterus and microphallus. Ovotestes were removed surgically. Vaginoplasty and clitoral recession were performed subsequently. Shortly after starting estrogen treatment at age 12, the patient expressed resistance to hormonal therapy because she was uncertain about her gender identity. SUMMARY: In all three cases, gender assignment was accepted by the patients' parents, but not by the patients themselves. These three cases inidicate that the traditional approach to gender assignment in newborns with AG may be associated with lack of acceptance on the patient's part, thus placing patients at risk for unwanted and unnecessary surgery and/or hormonal therapy.