Women's desires to have a certain number of children - or more specifically, a certain number of sons and/or daughters - can provide strong motivation for their use of contraception. This report uses data on 3,907 married, fecund, and nonpregnant Korean women obtained from the 1974 Korean National Fertility Survey to examine the consistency between respondents' answers and their fertility preferences; the determinants of fertility preferences; and the relationship between fertility preferences and contraceptive use. To do this, the authors constructed a multiple linear regression model containing 36 variables - of which 34 were independent and two, desired future births and present contraceptive use, were dependent. The major determinants of fertility preference were different than those of contraceptive use. The 34 independent variables explained 53% of the variance in desire for future births. Of these variables, age, marital duration, number of living children, and number of living sons explained 91% of this variance, with the number of living children being the most important. Family type (the influence of elders), as well as residence, wife's education, and attitude toward abortion were other leading determinants of the desire to have another child. Fertility preference was found to be an important intermediate variable in determining women's use of contraception. Number of sons was the most important determinant of contraceptive use and of actual reproductive behavior, although the differences in contraceptive use were negligible after the birth of two sons. Altogether, the 36 variables explained 17% of the total variance in contraceptive use. Desire for future births contributed most to this explained variance, with education, residence or lifetime migration status, family planning program availability, and family type also being important. Surprisingly, factors thought likely to be important in a society like Korea - birth order, number of siblings, work experience, and traditional attitudes - were unrelated to both fertility preference and contraceptive use or were subsumed by other variables. The report is amply supported by tables and data; a 25-item list of references (1954-79) is appended.