The results of a national survey of critical-care pharmaceutical services are reported. A questionnaire was mailed on March 8, 1988, to 1500 of the 6800 U.S. hospitals with more than 100 beds. The questionnaire was divided into four sections: (1) critical-care pharmaceutical services; (2) critical-care clinical pharmacy activity; (3) critical-care pharmacy standards; and (4) general information on the hospital, including pharmaceutical services provided to patients who are not critically ill. A total of 613 usable questionnaires were returned, for a response rate of 41%. Respondents from 124 hospitals reported having a critical-care satellite pharmacy. Critical-care beds served by satellite pharmacies were more likely to have unit dose drug distribution services than were critical-care beds not served by satellite pharmacies. The level of clinical and educational services provided to critical-care patients was significantly lower for intensive-care unit (ICU) pharmacists practicing in satellite pharmacies than it was for ICU pharmacists not associated with satellite pharmacies. There were no differences between satellite and nonsatellite ICU pharmacists in either the percentage of a typical day spent providing clinical services to critical-care patients or the percentage of critical-care patients receiving those services. Respondents expected their hospitals to add 301 full-time-equivalent critical-care pharmacists to their staffs during the next two years, for a 39% increase. Respondents strongly favored establishing 24-hour unit dose drug distribution and i.v. admixture services, providing inservice education to nurses and physicians, and providing written pharmacokinetic consultations.(ABSTRACT TRUNCATED AT 250 WORDS)