The authors conducted prenatal interviews and behavior observation sessions with forty-six families in which the woman was pregnant. These families were again interviewed and observed seven days, one month, two months, and six months after the child was born. The authors found that the most significant variable predicting differences in maternal attachment was the length of separation of mother and infant: less separation was associated with greater attachment. The next most significant variable was birth experience, followed in importance by the length of labor and prenatal attitudes and expectations.