This article examines the longitudinal trend of depressive symptoms in the Netherlands, using large-scale national data recorded over the period 1975–1996. Our analyses showed fluctuations in the overall longitudinal trend. On the basis of a general theoretical framework, we formulated hypotheses concerning which socio-demographic characteristics determine the likelihood of suffering from depressive symptoms and how these associations might have changed over time. Our results revealed that people on low incomes, unemployed people, unmarried people and those who had given up their church membership were associated with depressive symptoms. Some associations between socio-demographic categories and depressive symptoms have changed over time. Divorced people have become progressively less likely to suffer from depressive symptoms compared with married people, whereas the reverse holds for those who were never married. People on low incomes have become more likely to suffer from depressive symptoms over time in comparison to people with the highest incomes. Gender differences in these associations were also found: educational level and church attendance were more beneficial to women in protecting them from depressive symptoms than they were to men