Understanding the factors that influence an adolescent's decision to have sexual intercourse has important implications for theory, policy and practice. Because sexual intercourse is the single most important determinant of pregnancy, a fuller understanding of the factors related to heterosexual sexual activity is central to addressing the issue of teenage pregnancy. In addition, the current AIDS epidemic, along with the proliferation of other diseases that can be transmitted through sexual contact, make early sexual intercourse a potentially serious health hazard. Over the past two decades there has been a steady increase in the number of teenagers who are sexually active (Moore, 1992). In 1988, 52% of females aged from 15 to 19 had had premarital sexual intercourse, compared to only 29% in 1970. The birth rate among 15- to 17-year-old teens increased by 19% from 1986 to 1989. In addition, of the 12 million cases of sexually transmitted diseases that are estimated to occur annually, adolescents account for one quarter of those infected (Moore, 1992). In recent years a growing body of research has reported on the antecedents and correlates of sexual intercourse during the adolescent years. Numerous factors have been identified as being related to heterosexual adolescent sexual activity. They include poor school performance and low educational aspirations, alcohol and marijuana use, early pubertal development, low SES, living in a single-parent family, low religiosity, frequent dating with the same partner, and having a mother who was an early childbearer (Brooks-Gunn & Furstenberg, 1989; Flick, 1986; Hayes, 1987). Other factors that have been less extensively studied but have been linked to early sexual activity are a history of sexual abuse (Butler & Burton, 1990), a lack of positive experiences in school (Quinton & Rutter, 1988), poor neighborhood characteristics (Moore, Peterson, & Furstenberg, 1986), and a perception of poor employment prospects (Dash, 1989). The relationship between some variables and early heterosexual sexual activity remains ambiguous. For example, parental supervision was associated with less sexual activity in one study (Hogan & Kitagawa, 1984) but was unrelated in others (Inazu & Fox, 1980; Newcomer & Udry, 1983). Likewise, evidence regarding the degree to which peers influence sexual behavior is unclear (Hayes, 1987). There is some evidence that suggests that teens are more likely to be sexually active if they perceive that many of their peers are also sexually active (Flick, 1986). While there have been many studies that have examined disparate factors related to early sexual activity, fewer studies have integrated these multiple factors into a unified understanding of the causes and correlates of adolescent sexual intercourse. Exceptions to this include the work of Day (1992) and Udry and Billy (1987). In contrast to these past multivariate studies, the present investigation employs a cumulative risk approach and uses Bronfenbrenner's (1979, 1989) ecological perspective as a framework for organizing the risk factors that are examined. Borrowing from the field of epidemiology, a number of behavioral scientists have begun to employ a risk factor approach to the study of such diverse social problems as drug use (Newcomb, Maddahian, & Bentler, 1986), psychopathology (Rutter, 1978), and low intellectual achievement (Sameroff, Seifer, Barocas, Zax, & Greenspan, 1987). This approach suggests that there are probably many diverse paths to the development of a particular behavior (Newcomb, Maddahian, & Bentler, 1986) and that efforts to find a single cause may not be useful because most behaviors have multiple causes. Werner and Smith (1982) define risk factors as individual or environmental hazards that increase an individual's vulnerability to negative developmental outcomes. The presence of risk factors does not guarantee a negative outcome will occur, but simply increases the probability of its occurrence. …