SummaryMedical records from maternity clinics in three Norwegian cities, Oslo, Bergen and Trondheim, have been used to study the relationship between lactation and post-partum amenorrhoea during the period 1860–1964. Resumption of menses after a period of post-partum amenorrhoea has been examined separately during lactation, after cessation of lactation, and in the absence of lactation, in 5250 cases. During ongoing lactation the median duration of post-partum amenorrhoea was found by survival analysis to be 12 months for women who had given birth before 1900 and 6 months for women who had given birth after 1900. In a Cox regression analysis, age of menarche and parity were found to influence the duration of post-partum amenorrhoea, in addition to the child's birth year. With a later age of menarche there was longer amenorrhoea, and primiparae resumed menses earlier than multiparae. After cessation of breast-feeding the median duration of amenorrhoea was 1 month, the only variable significantly influencing it being the age of menarche, a younger age showing a shorter duration. In the absence of lactation the survivor function started to decline after 3–4 weeks, and had a half-life of 2 months.None of the variables indicating the woman's social condition seemed to influence the duration of amenorrhoea. The decline in the duration of post-partum amenorrhoea during ongoing lactation from 1860 until today is therefore probably caused by a changed breast-feeding pattern rather than by improved nutrition for women giving birth.