Background: In the 27 nursing homes located within the Veterans Affairs (VA) Central Region, the proportion of residents with pressure ulcers in 1986 varied from 2% to 16%. Three of these nursing homes were selected for study: nursing home A from the highest prevalence quintile, B from close to the median, and C from the lowest quintile. Methods: Two indicators of pressure ulcer occurrence were calculated: the point-in-time prevalence of bedsores and the percentage of residents who were free of pressure ulcers at the beginning of a six-month study period but who had bedsores six months later. Data were also collected in each nursing home to determine the bedsore status at the time of admission and the resident’s location if and when a bedsore began. Results: The three institutions were generally similar in available measures of casemix and severity of illness. The ratio of nursing and medical personnel to residents was 29%–76% lower in nursing home A than in B or C. In nursing home A, the turnover of nursing personnel was about twice as rapid as that in B or C. In each institution the pressure ulcer statistics showed little variation from one six-month period to another. The average rates in nursing home A were 15.3% for prevalence and 10.3% for the six-month conversion from bedsore-negative to bedsore-positive status. The average rates in facilities B and C were, respectively, 6.9% and 3.5% for prevalence and 4.7% and 4.2% for the six-month conversion from negative to positive status. Furthermore, the number of new bedsores that developed during uninterrupted nursing home residence, per 100,000 resident days, was 36.5 in A, 10.8 in B, and 2.1 in C. Conclusions: The pressure ulcer statistics in nursing homes B and C were consistently superior to those in A. The interinstitutional differences could not be explained by the comparisons of scoring methodologies and of casemixes that were made. It is hypothesized that more favorable staffing patterns in B and C than in A contributed to more effective prevention of bedsores in the former two institutions.