Data-gathering techniques included self-administered questionnaires to all staff, group and self-administered questionnaires to all girls, semi-structured interviews with key executives and a randomly selected sample of girls, and casual conversations and unstructured interviews with individual staff and youths. Other techniques were conversations with groups of youths, and casual and systematic observation, often while attending various recreational, educational, and therapeutic programs at the institutions. Questionnaire responses provided the main information source. A total of 396 girls (93 percent) and 262 staff (about 75 percent) completed both questionnaires. Overall rates for each of the specific indicators of homosexual activity were 14 percent for 'going with' or 'being married,' 10 percent for passionate kissing, 10 percent for writing love letters, and 7 percent for sex beyond hugging and kissing. The percentage reporting at least one of the aforementioned homosexual experiences was 17 percent. The rate varied from 6 percent to 30 percent, depending on the institution. Homosexuality was equally prevalent in the coed and all-female institutions. One probable reason is that inmates in both single-sexed and coed prisons have much more opportunity for interacting with same-sexed peers. Prior homosexuality was found to be far more important and deprived conditions of confinement less important in explaining institutional homosexuality than they were previously assumed to be. The impact, prevalence, and causes of prior homosexuality require further investigations. Data suggest that some aspect of incarceration contributes both to prior and current homosexuality. Tabular data, notes, and 25 references are provided. For related documents, see NCJ-70361-68 and 70370-77.