Objective: To review the activities in selected developed countries fo
r strategies to identify maternal deaths, the impact of these strategi
es on underreporting, and the information needed to understand the eve
nts leading to death. Data Sources: We reviewed the literature from th
e United States, Europe, and Australia for publications dealing with m
aternal death identification and investigation from 1980 to April 1995
. We also obtained information directly from researchers involved in m
ajor maternal mortality studies. Methods of Study Selection: We includ
ed all 31 reports (from 14 countries) that discussed methods to improv
e the ascertainment of maternal deaths beyond the routine use of vital
registration. Because of the nature of the subject matter, almost all
reports relied on descriptive epidemiology. Data Extraction and Synth
esis: We found that a variety of methods can be used to improve the as
certainment of maternal deaths, including linkage of birth and fetal d
eath certificates, check-boxes on death certificates, periodic review
of deaths of reproductive-age women, and ongoing birth registries and
medical audits. Information from a variety of sources is also needed t
o understand the events leading to death. Conclusion: The numbers of d
eaths due to pregnancy and its complications are underestimated in mos
t developed countries. Improved ascertainment of maternal death is nee
ded to determine the magnitude of the problem and to assess trends and
identify risk groups, allowing development of appropriate and effecti
ve strategies to prevent the morbidity and mortality associated with p
regnancy.