We analyzed the impact of the Medicare price for outpatient maintenance dialysis on the level and composition of staffing of dialysis units, and on the length (duration) of the average hemodialysis treatment for the period 1982 to 1987. With higher prices, dialysis units employ more total staff hours per patient per week (P less than 0.01) and use a higher proportion of registered nurses (RN) compared with other staff such as technicians and licensed practical nurses (LPN). Higher Medicare prices are also associated with longer average hemodialysis treatment duration for freestanding units (P less than 0.02), but there was no statistically significant association for hospital units. The impact of the Medicare price on both staffing and average treatment duration takes into account the calendar time trend, ie, 1987 versus 1982, which has a separate impact from the price effect. For example, to avoid the complication of technology changes over calendar time, we considered the impact of the price change at one point in time. Freestanding units that had a price cut of more than +15 in 1983, reported twice the reduction in the average dialysis treatment duration (-0.38 hours) compared with -0.20 hours (P = 0.01) for units that had a cut of less than +15. An analysis of the impact of average treatment duration on patient mortality provided some evidence that shorter treatment duration in freestanding units is associated with higher mortality. This result needs further analysis. There was no statistically significant relationship between the level of staffing in the dialysis unit or the percent of staff who were RNs and 3-year patient survival. The analysis in this report was based on 1,213 and 1,058 data forms filed with Medicare by freestanding and hospital units, respectively, for 1982 and 1987. In addition, we used a Cox proportional hazards model (multivariate) for the analysis of the 3-year patient survival for 14,807 new hemodialysis patients who started dialysis in 1984.