This historic prospective study assessed the relationship between dialyzer reuse practices and hemodialysis patient mortality through 1 year of follow-up. Medicare patient demographic and survival data were combined with dialyzer reuse data from the Centers for Disease Control and Prevention's annual survey of dialysis-related diseases. Data were analyzed for the US Medicare hemodialysis population of never transplanted patients prevalent on January 1, 1989, and January 1, 1990, who were treated in freestanding dialysis units that used primarily conventional (not high-flux) dialyzers. Time to mortality, or transplant, and other censoring on December 31st of each year was regressed with proportional hazards models on patient, dialysis unit, and reuse measures. Age-, race-, and diagnosis-standardized mortality ratios for dialysis units were also regressed with weighted least squares techniques against dialysis unit and reuse measures. The results showed that patients treated in dialysis units that disinfected dialyzers with a peracetic acid, hydrogen peroxide, acetic acid mixture, or glutaraldehyde experienced higher mortality than patients treated in units that used formalin or in units that did not reuse dialyzers. The relative risk of mortality, compared with patients treated in nonreuse dialysis units, was 1.17 (P = 0.010) for glutaraldehyde and 1.13 (P < 0.001) for the peracetic acid mixture. The relative risk for formalin compared with the reference group of nonreuse was 1.06 (P = 0.088). With adjustment for several patient and dialysis unit characteristics, dialyzer reuse with certain germicides was associated with a significantly elevated mortality risk. This elevated risk, the etiology of which is currently not known, may represent a large number of potentially avoidable deaths per year. Only a large, nationally based analysis of this type has sufficient sample size to detect mortality risks such as these.