Between 1978 and 1995, a ratio of 6.6:1 of boys to girls was seen in a specialty gender clinic for gender identity disorder. This paper evaluated several hypotheses for this phenomenon. The participants did not differ on demographic variables, nor on indicators of behavior problems. Although the percentage of boys and girls who met the criteria for GID was comparable, girls showed more extreme cross gender behavior. The paper concluded that external social factors were partly accountable for this discrepancy, as girls required having more extreme cross gender behavior than boys in order to be referred to the clinic.