Hospitalization tables of the U.S. Renal Data System allow description of national hospitalization rates among incident and prevalent dialysis patients in five-year age groups. These rates are further stratified by sex, race, and four primary disease categories. Based on these tables derived from the data on over 250,000 incident and prevalent patients during 1991 to 1993, a methodology is described that allows comparison of local (for example, dialysis facility) or regional "first admission" rates among incident and prevalent dialysis patients to the national rates. A standardized hospitalization ratio is introduced to facilitate such comparisons, and methods for assessing statistical significance are discussed. Since this methodology allows adjustment for age, race, sex, and primary disease, it can serve as useful tool for dialysis research. It can also be used at the dialysis facility level, alone or in conjunction with the standardized mortality ratio, to facilitate local quality assurance.