The empirical basis for theories and common wisdom regarding how to improve appropriate use of research evidence in policy decisions is unclear. One source of empirical evidence is interview studies with policy-makers. The aim of this systematic review was to summarise the evidence from interview studies of facilitators of, and barriers to, the use of research evidence by health policy-makers. We searched multiple databases, including Medline, Embase, Sociofile, PsychLit, PAIS, IBSS, IPSA and HealthStar in June 2000, hand-searched key journals and personally contacted investigators. We included interview studies with health policy-makers that covered their perceptions of the use of research evidence in health policy decisions at a national, regional or organisational level. Two reviewers independently assessed the relevance of retrieved articles, described the methods of included studies and extracted data that were summarised in tables and analysed qualitatively. We identified 24 studies that met our inclusion criteria. These studies included a total of 2041 interviews with health policy-makers. Assessments of the use of evidence were largely descriptive and qualitative, focusing on hypothetical scenarios or retrospective perceptions of the use of evidence in relation to specific cases. Perceived facilitators of, and barriers to, the use of evidence varied. The most commonly reported facilitators were personal contact (13/24), timely relevance (13/24), and the inclusion of summaries with policy recommendations (11/24). The most commonly reported barriers were absence of personal contact (11/24), lack of timeliness or relevance of research (9/24), mutual mistrust (8/24) and power and budget struggles (7/24). Interview studies with health policy-makers provide only limited support for commonly held beliefs about facilitators of, and barriers to, their use of evidence, and raise questions about commonsense proposals for improving the use of research for policy decisions. Two-way personal communication, the most common suggestion, may improve the appropriate use of research evidence, but it might also promote selective (inappropriate) use of research evidence.