Formative research is often used to inform intervention design, but the design process is rarely reported. This study describes how an integrated home visit intervention for newborns in Ghana was designed. As a first step in the design process, the known intervention parameters were listed, information required to refine the intervention was then identified and a formative research strategy designed. The strategy included synthesizing available data, collecting data on newborn care practices in homes and facilities, on barriers and facilitators to adopting desired behaviors and on practical issues such as whom to include in the intervention. The data were used to develop an intervention plan through workshops with national and international stakeholders and experts. The intervention plan was operationalized by district level committees. This included developing work plans, a creative brief for the materials and completing a community volunteer inventory. The intervention was then piloted and the intervention materials were finalized. The design process took over a year and was iterative. Throughout the process, literature was reviewed to identify the best practice. The intervention focuses on birth preparedness, using treated bednets in pregnancy, early and exclusive breastfeeding, thermal care, special care for small babies and prompt care seeking for newborns with danger signs. The need for a problem-solving approach was identified to help ensure behavior change. A subset of behaviors were already being performed adequately, or were the focus of other interventions, but were important to reinforce in the visits. These include attending antenatal care and care seeking for danger signs in pregnancy. On the basis of the intervention content, the timing of newborn deaths and the acceptability of visits, two antenatal and three visits in the first week of life (days 1, 3 and 7) were planned. Several household members were identified to include in the visits as they were involved in newborn care or they made financial decisions. Birth attendants and health workers were often the locus of control for immediate newborn care, and sensitization activities were designed to improve their practices and to help ensure that families received consistent messages. An existing cadre of community volunteers was identified to deliver the intervention-these volunteers were already trusted and accepted by the community, already visited pregnant women at home and had the time and commitment to deliver the intervention. A supervision and remuneration system was developed through discussions with the volunteers and at the planning workshops. The need for community entry activities was identified to garner community support for the intervention, to encourage self-identification of pregnant and delivered women and to motivate the volunteer through community recognition. Formative research is an essential step in helping to ensure the development of an effective, appropriate and sustainable intervention.