This thesis examines whether measurement of morbidity prevalence through survey methods provides a suitable alternative to mortality measurement for safe motherhood programme needs assessment. It considers the validity of a survey instrument by comparing results from a questionnaire on near-miss obstetric complications to hospital clinical data. Three groups of women -with severe obstetric complications, mild obstetric complications and with a normal delivery - were identified retrospectively in three hospitals in South Benin and interviewed at home using a questionnaire. The complications of interest were eclampsia, haemorrhage, dystocia and infections of the genital tract. The concept of near-miss death event was used to identify women with severe episodes of morbidity. The aim of the analysis was to find questions with very high specificity for measuring the prevalence of obstetric conditions even at the expense of sensitivity. The questionnaire was able to detect, with sufficient accuracy, eclamptic fits, abnormal bleeding in the third trimester for a recall period of at least 3-4 years, and all episodes of haemorrhage independent of timing within a shorter period of 2 years. The specificity of questions and combinations of questions for dystocia and infections of the genital tract was weak, and generated disappointing results except when information on treatment was included. Overall, better results were achieved for antepartum and acute events than complications defined as such because they are at the extreme end of a continuum. Severity only made a positive difference in the case of eclampsia with an increase in sensitivity. 1 These results are interpreted in the light of methodological constraints and findings from similar studies. Although the study could support the use of individual interview surveys for eclampsia and haemorrhage, this methodology cannot be readily recommended in view of the insufficient specificity reported elsewhere. The way forward in terms of morbidity information as well as the future of the near-miss concept is presented in the final chapter.