Numerous researchers have examined the incidence, correlates, and predictors of childlessness. Few, however, have examined changes in intended childlessness because the longitudinal data required to track these changes are rare. We utilize the National Survey of Families and Households to examine trends in intentions to remain childless. We include both demographic and ideational variables in the analysis, and we focus on respondents between the ages of 19 and 39 years who had not had children at the beginning of the study. The largest group wants children but still postpone childbearing. The next largest group carries out their intention to have children. The third largest group switches from wanting children to not wanting children. Some are consistently childless in both surveys. Finally, a relatively small group did not intend to have a child in the first survey but subsequently had a child. Marital status is the most salient predictor for having children, but cohabitors also are more likely to have children than are single noncohabitors. Rates of childlessness in the United States have varied substantially over the past several decades. Morgan (1991), for example, reports that census data show slightly over 15% of White women born in the mid-1880s remained childless. This childless rate increased to over 25% of women born in 1910 who reached normal childbearing age during the Depression. The percentage then dropped dramatically to about 10% of White women born in 1935 who reached childbearing age during the baby boom. Since then, the rate has increased again, with a projected childless rate of 22% for women born in 1962 (Morgan & Chen, 1992). Non-White women experienced a similar increase in childlessness during the Depression and a similar decline during the baby boom, but they have not participated in the post-baby boom rise in childlessness to the extent that White women have (Chen & Morgan, 1991.) These fluctuations indicate that potential parents do respond to economic and social conditions, even when modern and efficient contraception is not available. (See also May, 1995, and Friedman, Hechter, & Kanazawa, 1994.) Because demographers have been concerned with population growth arising from high fertility, their explanations for why people have children are designed to explain fertility decline. Such theories provide incomplete explanations for fertility decline during demographic transition (Mason, 1997) and are even less satisfactory in accounting for the persistence of childbearing in highly developed societies (Schoen, Young, Nathanson, Fields, & Astone, 1997). New approaches to understanding why people continue to have children have been suggested (Axinn & Thornton, 1996; Friedman et al., 1994; Rovi; 1994). Given recent increases in childlessness, these new approaches also should be able to account for the decision not to have children. We examine factors related to persistence and change in decisions to remain childless. The general trends in childlessness that we have noted combine both the infertile (or involuntarily childless) and the voluntarily childless. Though such trends reflect and affect the overall structure and composition of society, the rate of involuntary childlessness has declined as a result of better health and a general decrease in sterility caused by sexually transmitted diseases. Voluntary childlessness, on the other hand, reflects the choices of potential parents and has varied substantially by period and cohort. By all accounts, voluntary childlessness has increased in recent decades. We develop and test a model of persistence and change in voluntary childlessness. EXPLANATIONS FOR VOLUNTARY CHILDLESSNESS Prevailing theories of voluntary childlessness have tended to emphasize either a rational choice approach, focusing on the costs and benefits of having children, or ideational approaches, focusing more on values and norms. …