In reviewing the literature on burn therapy and observing clinical burn care, we noted differences among institutions and individual experts in several areas. To study variation in burn care, we surveyed the 140 burn centers listed by the American Burn Association to determine how burn care is currently administered in the United States and Canada. Responses were obtained from 83 hospitals (60%). The survey addressed resuscitation, operative and nonoperative wound care, medications, antimicrobial agents, and pain control. The major influence on care appeared to be the experience of the director (considered "very influential" in 85%) compared with the literature ("very influential" in 12%) and habit/what works for us ("very influential" in 48%). The Parkland formula was used "always" or "often" by 78%, and the Brooke formula "never" by 81% of respondents. Lactated Ringer's solution was the most popular initial fluid, and most (78%) respondents changed fluids after 24 hours. However, the fluids used in the second 24 hours varied equally among several choices. The use of colloids also varied without a set pattern. Furosemide (Lasix) and nonsteroidal antiinflammatory drugs were used "rarely" or "never" by 67% of centers in the acute stage. H2 blockers were used for gastritis prophylaxis "always" or "often" in 60% (vs 53% for antacids and 20% for sucralfate [Carafate]). Tube feedings were started on day 1 after burn injury "always" by less than 30% of centers. Total parenteral nutrition was not commonly used. Most centers use of silver sulfadiazine on the body and hands, but facial topical antimicrobial therapy varied.(ABSTRACT TRUNCATED AT 250 WORDS)