Advances in RRT in the last few years have resulted in multiple RRT modalities available for treating ARF in the ICU. CRRT is gaining greater acceptance with the use of venovenous access and its advantages in hemodynamically unstable patients. There are little data as to the best modality of RRT. There are few randomized controlled trials and most existing studies are retrospective and poorly controlled. Many confounders exist, such as severity of illness and etiology of renal failure, which are probably the most important factors affecting outcome in ICU patients with ARF. Some recent studies also suggest that higher doses of dialysis confer a survival advantage. Choice of modality should probably be tailored to the needs of the individual patient. IHD is best for patients requiring rapid metabolic control (eg, in hyperkalemia), whereas volume overload is best managed with CRRT. Patients who are hemodynamically unstable or who have increased intracranial pressure are best treated with CRRT. Patients in whom anticoagulation is contraindicated might be better managed with IHD unless CRRT with citrate is used. CRRT is limited by its greater cost and demands on nursing time and the constraint it places on a patient's mobility. Theoretically, the choice of RRT might also depend on the underlying disease and etiology of ARF, but this question requires further study. The choice of modality should be based on the clinical status of the patient, the resources available in the institution, and the cost of therapy. © 2005 Elsevier Inc. All rights reserved.