Dramatic improvements have been made in child survival over the last 30 years. However, despite the gains of the child survival revolution, neonatal mortality rates have stagnated as infant and child mortality has decreased. Every year 4 million newborns die before they reach 28 days of life. The Millennium Development Goals urge the reduction of under five mortality (from the 1990 level) by two thirds by the year 2015; this will not be achievable unless there is at least a halving in the number of neonatal deaths, which currently account for 36% of all childhood deaths. Some promising interventions have recently been shown to improve newborn survival through home and community based care and there is considerable interest in whether these can be implemented on a wide scale. They are particularly needed where a large proportion of births take place at home and access to health services is suboptimal. Such interventions depend on understanding sociocultural factors that form the basis for newborn care practices. Research elucidating these factors has recently been identified as a priority in several peer reviewed publications and within the international health community. Although the number of newborn deaths is highest in South Asia, the risk of newborn death is highest in Sub Saharan Africa, with West Africa having the highest rates. The present study critically examined the social, cultural, and behavioural factors that play a role in determining care practices during childbirth and the neonatal period in Kintampo District in rural Ghana. A qualitative, ethnographic, study design was used including participant observation, in-depth interviews, semi- structured interviews, expert interviews, narratives, and group discussions, with grounded theory as the guiding theoretical paradigm. It was carried out in four sites: Kintampo town and three villages, Apesika, Jema, and Kawampe. The study benefited from the ongoing ObaapaVitA Vitamin A maternal mortality trial database. This allowed triangulation of the ethnographic findings through analysis of birth cohort data on all 2,878 singletons born alive to mothers in the trial in Kintampo District within the year July 2003 - June 2004. Available data included: location of birth, presence of an attendant, wrapping and drying after birth, substances applied to the umbilical cord, bathing and early infant feeding practices. Narrative interviews from verbal autopsies conducted through the ObaapaVitA trial were also used to capture information on actual newborn deaths. The study findings are presented separately for the following three domains: Pregnancy and Preparation for Childbirth; Neonatal Care Practices; and Newborn Illness, Death and Care Seeking. These highlight several gaps in current practices where improvements might lead to reductions in neonatal mortality. Lessons learned have also been drawn together, both from the perspective of implications for the design of interventions to reduce neonatal mortality and concerning methodological issues in conducting formative research on newborn care practices.