The authors evaluated the potential of a variety of anesthetics in mice to produce subsequent alterations in host defenses. Specific monoclonal antibodies and immunofluorescent microscopy were used to enumerate splenic helper/inducer: suppressor/cytotoxic lymphocyte ratios (HSR), and resistance to bacterial challenge was evaluated by a cecal ligation and puncture (CLP) model. Two hours of anesthesia with the intravenous agents ketamine and pentobarbital and with the inhalational agents isoflurane, enflurane, halothane, and halothane-nitrous oxide, were utilized. All anesthetics produced marked depression in the HSR, measured 24 h postanesthesia (P less than 0.05); with all agents, helper T-cell populations were decreased and suppressor populations increased. The HSR remained depressed 72 h postanesthetic, following both ketamine and halothane anesthesia (P less than 0.05). A dose-response curve was determined with enflurane; increasing the anesthetic time from 1 to 6 h resulted in progressively greater depression of the HSR 24 h later. Changes in lymphocyte subtypes of similar magnitude were found in mice after burn injury or hind limb crush injury and amputation, whereas simple laparotomy did not produce such changes. Serum corticosterone levels were not elevated 24 h post-anesthetic with enflurane, suggesting that the alterations were not nonspecific stress reactions. Resistance to sepsis was determined by measuring survival for 96 h after CLP. With CLP performed 24 h following 2 h anesthesia, mortality was increased from normal: control mortality 36.3%; ketamine 65.0% (P less than 0.023); isoflurane 69.5% (P less than 0.006); enflurane 84.2% (P less than 0.0002).(ABSTRACT TRUNCATED AT 250 WORDS)