BACKGROUND: TB and HIV form a deadly synergy in much of the developing world, especially Africa. Interventions to reduce the impact of these diseases at community level are urgently needed. This paper presents the design of a community randomised trial to evaluate the impact of two complex interventions on the prevalence of tuberculosis (TB) in high HIV prevalence settings in Zambia and South Africa. METHODS: The interaction between TB and HIV is reviewed and possible interventions that could reduce the prevalence of TB in HIV-endemic populations are discussed. Two of these interventions are described in detail and the design of a 2 x 2 factorial community randomised trial to test these interventions is presented. The limitations and challenges of the design are identified and discussed. CONCLUSION: There is an urgent need to reduce the prevalence of TB in communities highly affected by HIV. Potential interventions are complex and require innovative trial designs to provide the rigorous evidence needed to inform health policy makers and to ensure that resources are used optimally. TRIAL REGISTRATION: Number: ISRCTN36729271.