Objective: To estimate the risk of having a low birth-weight infant associated with changes in social, environmental, and genetic factors. Design: Population based, historical cohort study using the Danish medical birth registry and Statistic Denmark's fertility database. Subjects: All women who had a low birthweight infant (<2500 g) (index birth) and a subsequent liveborn infant (outcome birth) in Denmark between 1980 and 1992 (exposed cohort, n = 11 069) and a random sample of the population who gave birth to an infant weighing greater than or equal to 2500 g and to a subsequent liveborn infant (unexposed cohort, n = 10 211), Main outcome measures: Risk of having a low birthweight infant in the outcome birth as a function of changes in male partner, area of residence, type of job, and social status between the two births. Results: Women in the exposed cohort showed a high risk (18.5%) of having a subsequent low birthweight infant while women in the unexposed cohort had a risk of 2.8%. After adjustment for initial social status, a decline in social statics increased the absolute risk of having a low birthweight infant by about 5% in both cohorts, though this was significant only in the unexposed cohort, Change of male partner did not modify the risk of low birth weight in either cohort. Conclusion: Having had a low birthweight infant and a decline in social status are strong risk factors for having a low birthweight infant subsequently.