This paper focuses on the evidence for the effectiveness of community-based outreach intervention as one component of a comprehensive HIV prevention model for preventing HIV infection in injecting drug user (IDU) populations. Three empirical questions guided the review of the evidence. This article includes primarily published literature on community-based outreach derived mostly from developing countries but also unpublished literature. Wherever possible, evidence from multi-country, multi-site studies or meta-analytical studies is included. More than 40 published studies reveal that injecting drug users (IDUs), who are reached by community-based outreach and provided with access to risk reduction services, report reducing HIV risk behaviours. The strength of the evidence was assessed using Hill’s criteria, which permit a review of multiple studies with different designs. Using the criteria, it is possible to infer causation about the evidence of effectiveness of the intervention. The evidence for the effectiveness of a community-based outreach strategy is strong. Despite evidence from 20 years of evaluation studies of the effectiveness of community-based outreach, a huge gap exists in most countries between the number of IDUs who want or could benefit from outreach services and the number of IDUs who actually receive them. Findings from evaluation studies on the effectiveness of community-based outreach must be made accessible, disseminated globally and provided to policy- and decision-makers to persuade them to take action and implement scaled-up prevention programmes. This requires ongoing advocacy and constant strengthening of the evidence base. Plans are needed to link evidence-based findings with technical assistance as well as training to enhance the capacity of regions and countries to introduce, scale up and sustain HIV prevention outreach to IDUs as part of a comprehensive HIV prevention strategy. © 2005 Elsevier B.V. All rights reserved.