This study seeks to examine the factors associated with neonatal mortality in Bangladesh based on the Bangladesh Demographic Health Survey 2004 data. It is observed that 54.4 per cent of the younger mothers (< 25 years) experienced neonatal mortality, while about 46 per cent mothers of age 25 years and above experienced neonatal mortality. Male children have a higher neonatal mortality rate as compared to female children. The analysis reveals that mothers’ age at birth is inversely related to neonatal mortality. Children born to mothers aged below 20 years are 1.5 times more expected to die in their neonatal period as compared to children born to mothers between 20–29 years. Children born less than two years after a previous birth are about 2 times more likely to die in their neonatal period as compared to children born after an interval of four years or more of the previous birth. The neonatal mortality is 1.8 times more for the first birth order compare to birth order 2–3. Sylhet division has the highest neonatal mortality rate. Children of illiterate mothers are more than 1.5 times likely to die before completing their first month of birth as compared to children born to mothers with education up to higher secondary or above. Again neonatal mortality is highest among the poorest asset groups, which is almost double as compared to households from the richest asset group. Multivariate logistic regression analysis showed that division, mother’s age at birth, mother’s education, previous birth interval, region, number of antenatal care visits, household asset index, received TT injections and place of delivery likely to affect neonatal deaths.