BACKGROUND: Adults with end-stage renal disease treated with dialysis experience a high burden of foot ulceration and lower extremity amputation. However, the risk factors for foot ulceration in the dialysis population are incompletely understood due to the lack of high-quality prospective evidence. This article outlines the design of a prospective observational cohort study, which aims to investigate the risk factors for foot ulceration in adults on dialysis. METHODS/DESIGN: This study will recruit 430 participants with end-stage renal disease on dialysis from satellite and home-therapy dialysis units across multiple health organisations in Melbourne, Victoria, Australia. Data collection at baseline will include a participant interview, medical record review, completion of a health-status questionnaire and a non-invasive foot assessment. Twenty participants will also be recruited to a reliability study to evaluate the reproducibility of testing procedures. Primary outcome data includes: new foot ulcer(s). Secondary outcome data includes: number of new foot ulcers, time to onset of new foot ulcer(s), new lower extremity amputation(s), episodes of infection of the foot or lower extremity, episodes of osteomyelitis, foot-related hospitalisations, revascularisation procedure(s) of the lower extremity, new podiatry interventions, kidney transplantation, and mortality. Participants will be assessed at baseline, and at 12 months they will be evaluated for the primary and secondary outcomes. Multivariate Cox proportional hazards models will be used to assess predictors of new foot ulceration and time to event secondary outcomes. Logistic regression will be used for binary outcomes including prevalence of foot ulcerations. DISCUSSION: This is the first multi-centre prospective observational cohort study to investigate risk factors for foot ulceration in adults with end-stage renal disease on dialysis. This study will improve on prior studies by using prospective methods, multi-centre recruitment, statistical methods to control for confounding variables, and a pre-specified sample size estimation. The findings can inform the design of future trials evaluating the effectiveness of clinical interventions, which may lead to improved patient outcomes in the dialysis setting.