This study aimed to explore researchers' and policy stakeholders' perceptions of the usefulness of public health research for policy, using the example of maternal health in Ghana. Sixty-nine government decisionmakers, maternal health policy stakeholders and researchers were interviewed. Concepts of research were broad. Research was dichotomised into `big', formal research and 'small', applied research such as operations research. 'Small research' was highly valued, due to its speedy completion and its focus on topics pertinent to service delivery; big research was not always considered necessary. Effectiveness research, one type of `big research', was not highly valued. Interviewees tended to feel that 'effective' policies and programmes could be designed once there was a thorough understanding of the situation. There was an implicit assumption that as long as these interventions were implemented well, they would be effective. Six dimensionso f local applicability/transferabilitwye re identified.T he most influential factors were the ease with which the intervention could be implemented, the study's congruence with interviewees' previous experiences and the perceived need for the intervention. Little attention was paid to study findings. Judgements of an intervention's potential effectiveness tended to be based on the ease of implementation or knowledge of similar projects. Adaptation was considered to be crucial, although often conceptualised not as a factor within local applicability/transferability assessments, but rather a distinct, essential step in the research use process. This study suggests that the factors of local applicability/transferability frequently cited in the literature do not reflect those considered to be most important by stakeholders in Ghana.