Background It is beyond doubt that blood transfusion services have added to the decline in maternal mortality in high-resource countries. To quantify the clinical benefit of red blood cell (RBC) transfusion in obstetric care, we performed a hypothetical experimental study using data from a prospective nationwide cohort of women giving birth in the Netherlands. Study Design and Methods Data were abstracted from a nationwide cohort study on severe maternal morbidity, including obstetric haemorrhage requiring 4 or more units of RBC, to obtain an observed and a hypothetical control group consisting of the same women. In the hypothetical control group, we simulated a situation where RBC transfusion was unavailable and estimated how many of these women would have died in that situation. A questionnaire survey asked experts in major (obstetric) haemorrhage to choose a critical minimal number of RBC transfusions at which a woman with obstetric haemorrhage would have died if RBC transfusion was not available. Maternal mortality rate per 100000 maternities [maternal mortality ratios (MMR)] and relative risk were calculated for the observed and hypothetical group. Results The observed MMR was 13 per 100000 maternities. According to 47 responding experts, the median number of RBC units without which a woman would have most probably died was nine, resulting in a hypothetical MMR of 87per 100000 maternities (relative risk 6·5; 95% confidence interval 4·2-10·0). Conclusions It can be expected that unavailability of RBC transfusion in obstetric care increases the risk of maternal death 6.5-fold. Blood transfusion thus largely contributes to the decline of MMR and would also be an important pillar of improving quality of care in resource-poor settings. © 2012 International Society of Blood Transfusion.