BACKGROUND: Rapid diagnostic tests (RDTs) for malaria are at the early stages of introduction across malaria endemic countries. This is central to efforts to decrease malaria overdiagnosis and the consequent overuse of valuable anti-malarials and underdiagnosis of alternative causes of fever. Evidence of the effect of introducing RDTs on the overprescription of anti-malarials is mixed. A recent trial in rural health facilities in Ghana reduced overprescription of anti-malarials, but found that 45.5% patients who tested negative with RDTs were still prescribed an anti-malarial. METHODS: A qualitative study of this trial was conducted, using in-depth interviews with a purposive sample of health workers involved in the trial, ranging from those who continued to prescribe anti-malarials to most patients with negative RDT results to those who largely restricted anti-malarials to patients with positive RDT results. Interviews explored the experiences of using RDTs and their results amongst trial participants. RESULTS: Meanings of RDTs were constructed by health workers through participation with the tests themselves as well as through interactions with colleagues, patients and the research team. These different modes of participation with the tests and their results led to a change in practice for some health workers, and reinforced existing practice for others. Many of the characteristics of RDTs were found to be inherently conducive to change, but the limited support from purveyors, lack of system antecedents for change and limited system readiness for change were apparent in the analysis. CONCLUSIONS: When introduced with a limited supporting package, RDTs were variously interpreted and used, reflecting how health workers had learnt how to use RDT results through participation. To build confidence of health workers in the face of negative RDT results, a supporting package should include local preparation for the innovation; unambiguous guidelines; training in alternative causes of disease; regular support for health workers to meet as communities of practice; interventions that address negotiation of health worker-patient relationships and encourage self-reflection of practice; feedback systems for results of quality control of RDTs; feedback systems of the results of their practice with RDTs; and RDT augmentation such as a technical and/or clinical troubleshooting resource.