Developmental Psychology 1996, Vol. 32, No. 1,3-11 Copyright 1996 by the American Psychological Association, Inc. 0012-1649/96/J3.00 Do Parents Influence the Sexual Orientation of Their Children? Findings From a Longitudinal Study of Lesbian Families Susan Golombok and Fiona Tasker City University Findings are presented of a longitudinal study of the sexual orientation of adults who had been raised as children in lesbian families. Twenty-five children of lesbian mothers and a control group of 21 children of heterosexual single mothers were first seen at age 9.5 years on average, and again at age 23.5 years on average. Standardized interviews were used to obtain data on sexual orientation from the young adults in the follow-up study, and on family characteristics and children's gender role behavior from the mothers and their children in the initial study. Although those from lesbian fami- lies were more likely to explore same-sex relationships, particularly if their childhood family envi- ronment was characterized by an openness and acceptance of lesbian and gay relationships, the large majority of children who grew up in lesbian families identified as heterosexual. Opinion varies among biological and psychological theorists regarding the extent to which it is possible for parents to influ- ence the sexual orientation of their children. From a purely bi- ological perspective, parents should make little difference. In contrast, psychoanalytic theorists believe that relationships with parents in childhood are central to the development of sex- ual orientation in adult life. Research on adults raised in lesbian families provides an opportunity to test theoretical assumptions about the role of parents in their children's sexual orientation; if parents are influential in whether their children grow up to be heterosexual, lesbian, or gay, then it might be expected that lesbian parents would be more likely than heterosexual parents to have lesbian daughters and gay sons. With the exception of Gottman's (1990) investigation of adult daughters of lesbian mothers in which actual sexual behavior was not reported, re- search on lesbian families has focused on children rather than adults, and sexual orientation has not been assessed (Golombok, Spencer, & Rutter, 1983; Green, Mandel, Hotvedt, Gray,&Smith, 1986;Hoeffer, 1981; Kirkpatrick, Smith, & Roy, 1981; for a review, see Patterson, 1992). From the existing literature, it seems that no single factor de- termines whether a person will identify as heterosexual or ho- mosexual. The current view is that there are a variety of influ- ences, from the prenatal period onward, which may shape de- velopment in one direction or the other. Studies of gay men with twin brothers (Bailey & Pillard, 1991) and lesbian women with twin sisters (Bailey, Pillard, Neale, & Agyei, 1993) have found that a significantly greater proportion of monozygotic than di- Susan Golombok and Fiona Tasker, Family and Child Psychology Re- search Centre, City University, London, United Kingdom. We are grateful to Michael Rutter for his advice and encouragement at the early stages of the follow-up study and to Clare Murray for her help with coding data. We thank the Wellcome Trust (U.K.) for funding this research and the staff at the National Health Service Central Regis- ter (U.K.) for their assistance in tracing participants. Correspondence concerning this article should be addressed to Susan Golombok, Family and Child Psychology Research Centre, City Uni- versity, Northampton Square, London EC 1V OHB, United Kingdom. zygotic co-twins were gay or lesbian. The greater concordance between identical than nonidentical twin pairs indicates a ge- netic link to homosexuality, although this does not mean that a homosexual (or heterosexual) orientation is dependent on a specific genetic pattern. The identification of a genetic marker for male homosexuality has recently been reported by Hamer, Hu, Magnuson, Hu, and Pattatucci (1993). Of 40 pairs of brothers, both of whom were homosexual, 33 pairs were found to have a marker in a small region of the X chromosome, sug- gesting that there may be a specific gene, yet to be located, which is linked to male homosexuality. However, the presence of this gene, if it exists, would not necessarily determine a homosexual orientation, and not all homosexual men would necessarily pos- sess the gene (the marker was not found in 7 pairs of brothers). Instead, it may be one of many factors that influence develop- ment along a homosexual rather than a heterosexual course. Gonadal hormone levels may constitute another such factor. Although no consistent differences in gonadal hormone levels between heterosexual and homosexual adults have been identi- fied (Meyer-Bahlburg, 1984), there is evidence to suggest that the prenatal hormonal environment may play some part in the development of sexual orientation. Studies of women with con- genital adrenal hyperplasia (CAH), a genetically transmitted disorder in which malfunctioning adrenal glands produce high levels of androgens from the prenatal period onward, have found that these women were more likely to consider themselves to be bisexual or lesbian than were women who do not have the disorder, suggesting that raised levels of androgens prenatally may be associated with a lesbian sexual orientation (Dittman, Kappes, & Kappes, 1992; Money, Schwartz, & Lewis, 1984). In addition, a significantly greater proportion of women ex- posed in utero to the synthetic estrogen diethylstilbestrol (DES), an androgen derivative, reported bisexual or lesbian re- sponsiveness compared with both unexposed women from the same clinic and their unexposed sisters (Ehrhardt et al., 1985). It is important to note, however, that most of the women with CAH, and most of the women prenatally exposed to DES, were heterosexual despite their atypical endocrine history. On the basis of this research, together with animal research This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.