/. biosoc. Sci. (1995) 27, 179-192 INFANT AND CHILD MORTALITY LEVELS AND TRENDS IN BANGLADESH M. KABIR*, RAFIQUEL ISLAM CHOWDHURYf AND RUHUL AMINJ ^Department of Statistics and ^Computer and Information Technology Institute, Jaftangirnagar University, Savar, Dhaka, Bangladesh, and J University Linkage Program, Morgan State University, Baltimore, USA Summary. Infant and child mortality levels and trends in Bangladesh are examined using data from the 1989 Bangladesh Fertility Survey. Both infant and child mortality declined from the mid 1970s but infant mortality declined more quickly. The level of infant mortality in 1989 was around 100 per 1000 live births while child mortality (5q 0 ) was 200 per 1000 live births. Life table analysis confirms the change in infant and child mortality. The decline in infant mortality is attributed to the introduction of improved public health measures and access to maternal and child health services. Introduction Infant and child mortality (mortality during the first year of life and from age 1-4 years respectively) are important indicators for describing mortality conditions, health progress and the overall social and economic well-being of a country (UN, 1988). Infant and child mortality in Bangladesh are of interest to population researchers because of the apparent relationship with fertility and, indirectly, with the acceptance of modern contraception. The health benefits to mothers and their children resulting from the practice of contraception are a primary reason for providing family planning services. Preliminary analysis of Bangladesh maternity history data shows little change in infant mortality between 1979 and 1984. Erratic fluctuations in the levels may be attributed to sampling fluctuations but the underlying level for 1980-84 is about 130 per thousand births (Huq & Cleland, 1990). Following 1985, there is a sharp fall in infant mortality, which may be a genuine trend or reflect data errors. Child mortality, however, remains very high, though some decline is evident. This paper aims to establish whether the decline in child mortality is less than the decline in infant mortality. Data from the 1989 Bangladesh Fertility Survey will be compared with other sources using indirect methods; as these may be affected by errors of sampling, reporting or methodology, caution is advised. Direct estimates, based on maternity histories, may also be affected by sampling and reporting errors. 179