Trial design: Cluster randomised controlled trial. Objective: To compare current practice for encouraging homeless people to be screened for tuberculosis on a mobile digital X-ray unit in London, UK, with the additional use of volunteer peer educators who have direct experience of tuberculosis, homelessness or both. Participants: 46 hostels took part in the study, with a total of 2342 residents eligible for screening. The study took place between February 2012 and October 2013 at homeless hostels in London, UK. Intervention: At intervention sites, volunteer peer educators agreed to a work plan that involved moving around the hostel in conjunction with the hostel staff, and speaking to residents in order to encourage them to attend the screening. Randomisation: Cluster randomisation (by hostel) was performed using an internet-based service to ensure allocation concealment, with minimisation by hostel size and historical screening uptake. Blinding: Only the study statistician was blinded to the allocation of intervention or control arms. Primary outcome: The primary outcome was the number of eligible clients at a hostel venue screened for active pulmonary tuberculosis by the mobile X-ray unit. Results: A total of 59 hostels were considered for eligibility and 46 were randomised. Control sites had a total of 1192 residents, with a median uptake of 45% (IQR 3355). Intervention sites had 1150 eligible residents with a median uptake of 40% (IQR 25-61). Using Poisson regression to account for the clustered study design, hostel size and historical screening levels, there was no evidence that peer educators increased uptake (adjusted risk ratio 0.98; 95% CIs 0.80 to 1.20). The study team noted no adverse events. Conclusions: This study found no evidence that volunteer peer educators increased client uptake of mobile X-ray unit screening for tuberculosis. Further qualitative work should be undertaken to explore the possible ancillary benefits to peer volunteers.