Data from the Medical Birth Registry of Norway, covering more than one million births for the period 1967-1984, were used to study the magnitude and effects of selective fertility, which is the tendency for a woman to replace a perinatal loss. Variation in fertility after the first three births is studied, controlling for perinatal outcome of previous births, maternal age, and year of birth. Even after the first birth, fertility is higher after a perinatal loss. Selective fertility is more strongly present at each successive birth order, and at each birth order it is stronger among older women. As the average number of births per woman decreases, the force of selective fertility increases; that is, its importance has increased over time. Perinatal mortality at the third and fourth birth orders is particularly distorted by the mechanism of selective fertility in studies based on cross-sectional data. Mortality at second birth is exaggerated by 1%, at third birth by 8% to 20%, and at fourth birth by 18% to 27%, with the largest effects seen in the later periods. A major portion of the increase in perinatal mortality from the second to fourth birth seen in most studies based on cross-sectional data can be explained by the mechanism of selective fertility.