The Regionalization of Perinatal Care, an intervention study carried out in Tanjungsari, a subdistrict in rural West Java, aimed to develop a comprehensive maternal health program to improve maternal and perinatal health outcomes. The main inputs included training at all levels of the health care system (informal and formal) and the establishment of birthing homes in villages to make services more accessible. Special attention was given to referral, transportation, communication and appropriate case management, A social marketing program was conducted to inform people of the accessible birthing homes for clean delivery, located near the women, and with better transportation and communications to referral facilities should complications arise. The study design was longitudinal, following all pregnant women from early pregnancy until 42 days postpartum in an intervention and a comparison area. The population was +/- 90,000 in the intervention area and 40,000 in the comparison area. Inclusion criteria were all mother and infant units delivered between June 1st, 1992 and May 31st, 1993. Analysis showed the following results: Most women sought antenatal care (> 95%). In Tanjungsari, nearly 90% sought such care from professional providers as versus 75% in the control area of Cisalak. Most women with bleeding or bleeding and edema during pregnancy sought professional assistance in both the study and control areas. However, fever for more than 3 days received more attention in the study area versus control area (93 vs. 69%). Greater than 85% of deliveries in both areas were conducted by TBAs. However, in the study area, nearly one-third of those with intrapartum complications (17%) delivered in a health facility compared to one-tenth in the control area. This meant a hospital delivery, primarily with assistance of a doctor or doctor/midwife combination. Overall referral rates by TBAs were low -13% of women with complications in Tanjungsari and 6% in Cisalak. More women with intrapartum complications were referred in the study area than in the control, and more complied when referred. Women who suffered intrapartum complications were more likely to have a perinatal death. Perinatal deaths declined in Tanjungsari, but not significantly. However, the trend over the period of the intervention shows an improvement in the deliveries managed by TBAs with more deaths resulting in the hands of professionals. Either women were arriving too late or the quality of care could not meet the needs. There was no change in the levels or place of perinatal deaths in Cisalak.