Developmental Psychology 1995, Vol. 31, No. 1,31-42 Copyright 1995 by the American Psychological Association, Inc. Early Hormonal Influences on Childhood Sex-Typed Activity and Playmate Preferences: Implications for the Development of Sexual Orientation Sheri A. Berenbaum and Elizabeth Snyder University of Health Sciences/Chicago Medical School Early hormones have been hypothesized to influence both sexual orientation and related childhood sex-typed behaviors. This study examined hormonal influences on activity and playmate preferences in children with congenital adrenal hyperplasia (CAH) age 2.5 to 12 years (24 girls, 19 boys) and their relatives (16 girls, 25 boys). CAH girls preferred boys' toys and activities. They also had greater preference for boy playmates than controls, but this was largely attributable to 4 girls. CAH boys did not differ significantly from controls. Activity and playmate preferences were not related. It was concluded that early androgen has a major effect on childhood activity preferences but only a weak influence on playmate preferences; sex segregation does not arise from sex-typed activity prefer- ences; activity and playmate preferences may be differently predictive of sexual orientation; and hormonal influences on sexual orientation are likely to be complex. Studies of the development and causes of sex-typed behavior are important for understanding the development of sexual ori- entation, because these behaviors are predictive of adult sexual orientation, and the factors hypothesized to contribute to sex- typed behavior are often the same as those hypothesized to con- tribute to sexual orientation. Homosexual men and women have higher rates of sex-atypical behavior in childh'ood and adulthood than same-sex heterosexuals. Early hormones have been suggested to influence both sex-typed behavior and sexual orientation. Sheri A. Berenbaum and Elizabeth Snyder, Department of Psychol- ogy, University of Health Sciences/Chicago Medical School. This study was supported by National Institutes of Health Grant HD19644. We thank the following people who contributed to this proj- ect: Stephen Duck, Orville Green, David Klein, Ora Pescovitz, Gail Richards, and Julio Santiago generously provided access to their pa- tients and answered medical questions; Brenda Henderson, Barbara Johnson, Kim Ketterling, and Robyn Reed helped with data collection, scoring, and entry; Brenda Henderson helped to design and construct the playmate preference test; Jo-Anne Finegan suggested the use of a toy preference task; Kristina Korman and Valerie Manter-Kapanke as- sisted in data scoring, analysis, and manuscript preparation; and Mi- chael Bailey provided the results of unpublished analyses, helpful com- ments on the manuscript, and thoughtful and stimulating discussions about the nature and causes of behavioral sex differences. We are very grateful to the participants and their parents for their continuing coop- eration and enthusiastic participation in the study. Portions of these data were presented at the Third Annual Meeting of the American Psychological Society, Washington, DC, in June 1991 and at the biennial meeting of the Society for Research in Child Develop- ment, New Orleans, Louisiana, in March 1993. Correspondence concerning this article should be addressed to Sheri A. Berenbaum, Department of Psychology, University of Health Sci- ences/Chicago Medical School, 3333 Green Bay Road, North Chicago, Illinois 60064. Electronic mail may be sent via Internet to CMSBERENBAUM@MINNA.ACC.IIT.EDU. The largest differences between homosexual and heterosexual individuals appear on sex-typed activities and interests (see Bai- ley & Zucker, 1995, for a review of these studies). Both pro- spective and retrospective studies of childhood gender role be- havior reveal homosexual males to be more likely than hetero- sexual males to participate in girls' games, to be called a sissy, to experiment with cosmetics and jewelry, and to be less likely than heterosexual males to participate in sports. Similarly, les- bians retrospectively report that they were more likely than het- erosexual females to participate in sports, to be called a tomboy, and to be less likely than heterosexual females to experiment with cosmetics and jewelry and to wear dresses. Homosexual men and women have also reported that they played with oppo- site-sex playmates in childhood more than did heterosexual men and women (Bailey & Zucker, 1995). Differences between homosexual and heterosexual individu- als may not be confined to childhood activities. For example, heterosexual men have been reported to have higher spatial abil- ity and lower verbal ability than both homosexual men and het- erosexual women (Gladue, Beatty, Larson, & Staton, 1990; McCormick & Witelson, 1991). Both homosexual men and women may also have higher rates of left-handedness than het- erosexual comparisons (McCormick, Witelson, & Kingstone, 1990). It is important to note, however, that many other sex-typed behaviors are not related to sexual orientation. Most homosex- ual men and women have typical gender identity (Bailey & Zucker, 1995; Green, 1987; LeVay, 1993) and are similar to same-sex heterosexual individuals on aspects of mating psy- chology related to sexual selection that show sex differences, for example, interest in uncommitted sex (Bailey, Gaulin, Agyei, & Gladue, 1994). These comparisons are useful because sexual orientation differences that parallel sex differences suggest that some of the factors that cause sex differences also cause differences between homosexual and heterosexual individuals of the same sex. (Be- 31