The results of a 1985 national mail survey of pharmaceutical services in short-term hospitals are reported and compared with similar surveys conducted in 1975, 1978, and 1982. A sample of 809 hospitals was selected randomly from among the estimated 5480 U.S. short-term hospitals that employ a pharmacist on at least a part-time basis. The survey had an 84.6% response. Fewer than 10% of respondents did not have and were not planning for unit dose systems. The percentage of hospitals with complete or partial pharmacy-based i.v. admixture services increased since 1982 from 67.4% to 80.6%. More than half the respondents said floor-stock i.v. solutions were distributed by a department other than pharmacy. Only about half the respondents had both unit dose drug distribution and i.v. admixture services. Prime vendors were used by 87% of small hospitals, 75% of medium hospitals, and 51% of large hospitals. The percentage of pharmacies involved in patient monitoring increased from 27% in 1982 to 44% in 1985. The survey also collected data on other clinical services, computerization, inventory turnover, use of prime vendors, and selection of therapeutic alternates. Drug distribution services are improving, but the distribution of injectable drugs often is not under pharmacy control. Only 10% of respondents had comprehensive pharmaceutical services, defined as complete unit dose drug distribution and i.v. admixture services plus three or more clinical services.