Maternal morbidity and mortality are high in the Indian context, but the majority of maternal deaths could be avoided by prompt and effective access to intrapartum care (WHO, 1999). Understanding the care seeking responses to intrapartum morbidities is crucial if maternal health is to be effectively improved, and maternal mortality reduced. This paper presents the results of a prospective study of 388 women followed through delivery and traditional postpartum in rural Karnataka in southern India. In this setting, few women use the existing health facilities and most deliveries occur at home. The analysis uses quantitative data, collected via questionnaires administered to women both during pregnancy and immediately after delivery. By virtue of its prospective design, the study gives a unique insight into intentions for intrapartum care during pregnancy as well as events following morbidities during labour. Routine care in the intrapartum period, both within institutions and at home, and impediments to appropriate care are also examined. The study was designed to collect information about health seeking decisions made by women and their families as pregnancies unfolded, rather than trying to capture women’s experience from a retrospective instrument. The data set is therefore a rich source of quantitative information, which incorporates details of event sequences and health service utilization not previously collected in a Safe Motherhood study. Additional qualitative information was also available from concurrent in-depth interviews with pregnant women, their families, health care providers and other key informants in the area. The level of unplanned institutional care seeking during the intrapartum period within the study area was very high, increasing from 11% planning deliveries at a facility to an eventual 35% actually delivering in hospitals. In addition there was a significant move away from planned deliveries with the auxiliary nurse midwive (ANM), to births with a lay attendant or dai. The proportion of women who planned for an ANM to assist was 49%, as compared with the actual occurrence, which was less than half of this proportion. Perceived quality of care was found to be an important factor in health seeking behaviour, as was wealth, caste, education and experience of previous problems in pregnancy. Actual care given by a range of practitioners was found to contain both beneficial and undesirable elements. As a response to serious morbidities experienced within the study period, many women were able to seek care although sometimes after a long delay. Those women who experienced inadequate progression of labour pains were most likely to proceed unexpectedly to a hospital delivery.