Background
No previous study has provided national estimates of the prevalence of primary and secondary infertility in sizeable areas of sub-Saharan Africa.

Methods
Primary infertility is measured by the proportion childless among women who entered their first marriage at least 7 years before date of censoring. Secondary infertility is measured by the ‘subsequently infertile estimator’ from parous ever-married women. Exposure begins at the age of the woman at the birth of her first child, and exposure ends when the woman is of an age, which is 5 years lower than her age at censoring. These last 5 years are used to determine her status as infertile or fertile at the last observation 5 years before censoring. A woman is considered infertile at last observation if she has had no livebirths during the last 5 years before censoring, otherwise she is considered fertile. A woman who has not given birth at age a or later is defined as being ‘infertile subsequent to age a’. The index of the proportion subsequently infertile at age a is estimated as the number of women infertile subsequent to age a, divided by the total number of women observed at that age. Infertility is estimated for women age 20–44.

Results
Primary infertility is relatively low and it exceeds 3% in less than a third of the 28 African countries analysed. In contrast, elevated levels of secondary infertility prevail in most countries. Secondary infertility for women age 20–44 ranges from 5% in Togo to 23% in Central African Republic.

Conclusions
It is feasible to gauge national levels of primary and secondary infertility from population based surveys including a birth history. The prevalence of infertility of pathological origin is so high in sub-Saharan Africa that infertility is not merely an individual concern, it is a public health problem.