The prevalence of peritoneal dialysis (PD) in the USA is 12%. PD is less costly and as effective as hemodialysis (HD) for most patients. OBJECTIVE: To project savings from a shift in modality distribution (MD) to home-based therapies. METHOD: We developed a decision model to assess the cost impact of changes in MD, modality cost (MC) or numbers of patients. IPD was excluded. Using 1999 USRDS data, we calculated MC and applied patient numbers and MD. We previously surveyed US nephrologists who opined optimal MD to maximize survival was: hospital HD 9.2%; community full-care HD 33.9%; community self-care HD 13.8%; home HD 12.1%; CAPD 17.5%; APD 14.3%; and IPD 1.5%. To maximize cost effectiveness the percentages changed to 4.2, 24.9, 15.5, 16.5, 26.7, 12.4, and 1.0% respectively. The estimated total dialysis spend, excluding epoietin and non-ambulance transportation, was generated and compared to the total cost using the "optimized" MD. RESULTS: A total dialysis spend of $11.25 billion was estimated using MD from the USRDS. The total savings and additional patients who could be treated at the MD opined for maximum survival was $750 million (6.7%) and 15,745, respectively; for maximum cost savings, $1.06 billion (9.5%) and 22,977. Survey limitations, costing assumptions and modeling suggest caution that the potential savings identified may or may not be completely achievable. CONCLUSION: Effecting a change in MD in accordance with what US nephrologists reported to be acceptable could lead to substantial Medicare savings. Strategies to increase utilization of home and self care therapies, which might include earlier referral and empowerment of patient choice, should be considered.