Objective To evaluate maternity care, the feasibility of collecting the relevant data from hospital records, and to assess the usefulness for district health planning in the Mtwara Region, Tanzania.

Method A prospective study running over 2 years using the unmet obstetric need concept in all four hospitals in the region, covering all women undergoing major obstetric interventions. Data on indications, interventions, and pregnancy outcome were collected. Rates of major obstetric interventions performed for absolute maternal indications among all expected births in the study area were calculated to assess met and unmet need.

Results Of 2404 interventions, 91% were caesarean sections, most commonly indicated by cephalo-pelvic disproportion (46%). The rate of major obstetric intervention for absolute maternal indications per 100 expected births was 1.8% overall with wide rural–urban disparities (1.4% vs. 3.3%). Based on the estimated need of at least 2% of births requiring of a major obstetric intervention, seven of 20 divisions reached or surpassed this threshold, seven had a level between 1% and 2% and six did not reach 1%. Data were collected as part of the regular activities of the health staff, but external resources were needed for data management and analysis.

Conclusion We found a major unmet obstetric need in rural areas with a clear trend towards fewer interventions with longer distance from a hospital. Geographical mapping of the met obstetric need provided a detailed picture of deficiencies in the provision of maternity care and could be useful for priority setting and monitoring.