Data was obtained on Swedish women from a fertility register with linked census and vital statistics data. Cumulative and age specific rates of first birth were estimated and graphed. Graphs of age specific births clearly discern the pattern of baby boom in the mid-1960s followed by a 10-year period of decline, particularly among those 16-25 years old, and a shift to births at higher ages before 1984. Births increased for 25-35 year olds. 60% of women born in 1960 were childless at 25 years old, compared to 40% for women born in the 1940s. The timing of first births among cohorts with completed fertility was almost the same. Delayed first births became apparent only with examination of cohorts classified by educational level. For example, 12% of the 1940 cohort of women 20 years old with fewer years of schooling were childless, and 17% of the same cohort with more years of schooling were childless. Age specific risk curves showed women with lengthy educations to have low birth rates at young ages, rapid increases at around 25 years old, plateaus at around 30 years old, and increases at 35 years old. The risk of a fecund woman remaining childless and the probability of first birth was estimated. The actual and estimated values of childlessness were close. The largest differences were among women born during 1943-45 with lengthy educations. Childlessness was estimated to be 17-18% for later generations born after 1959. The prediction of first birth rates with a new method proposed by Martinelle and applied to first birth rates for women born in 1935 was quite close to actual rates; the proportional error was 0.052.