The purpose of this study was to develop causal models relative to the premarital use of contraception among male and female college students. Eleven predictor variables were incorporated in separate male and female models. Self~administered questionnaires were mailed to a random sample of dormitory residents at a large southeastern university. A path analysis procedure was performed on the data from those students who were nonvirgins. The male model accounted for 20 percent of the total variance. The most influential predictor variable was the frequency in which males engaged in coitus. The more frequently they engaged in coitus, the more likely they or their partner used a reliable contraceptive at their latest coitus. Males who had expected coitus to occur before it did were more likely to have used a reliable contraceptive than males who did not expect coitus to occur. This was the second most influential variable in the male model. The female model accounted for 15 percent of the total variance. The most influential predictor variable was the degree of emotional attachment the females had toward their sex partner. The greater the degree of emotional attachment, the greater the likelihood they or their partner used a reliable contraceptive at their latest coital experience. Females were also more likely to have used reliable contraception at their most recent sexual intercourse the greater the number of close friends who were thought to use contraception. This was the second most influential variable in the model. Methodological and theoretical implications were discussed and recommendations for future research were made. / Ph. D.