Three matched groups of predominantly homosexual men were compared using Weinberg and Williams' questionnaire. Two of the groups contained men who were currently heterosexually married and men who had been heterosexually married; the third group was unmarried controls. Reasons for marriage were reviewed and found to be a combination of situational factors and, to a greater extent, of the expectations of those men who married of a more negative societal reaction against homosexuality. Implications of these findings for therapy are briefly discussed.