Results are presented from an analysis of the intervals between births from Imesi, Western Nigeria. The mean interval following births of children who survived to 1 year was 35·03 months, but nearly half this value at 18·8 months following stillbirths or deaths under 1 year. Both these results confirm an early study done in the village (Martin, Morley & Woodland, 1964), and the role of post-partum sexual abstention as an explanation for these results is discussed. Attempts were made to look for effects of birth interval variation on the survival of the following child, but no strong trends emerged. Similarly, analyses of many growth variables by preceding and following birth intervals gave very little evidence of real effects. There were, however, hints of trends which were possibly caused by birth order biases; higher birth orders had been found to have correspondingly higher birth intervals. An intriguing result was that those children who had ever suffered from kwashiorkor showed a significantly higher preceding mean birth interval, and also a higher mean interval to the next birth.