Humidified soda lime is commonly used to eliminate carbon dioxide from the circulatory system. Little is known about adverse reactions to accidentally dried soda lime. Therefore, a case of unexpected absorption of halothane by dry soda lime is reported. These observations were confirmed by a simulation with relevance to anaesthetic practice. CASE REPORT. A 46-year-old ASA class I patient was scheduled for elective surgery. After induction of general anaesthesia with 500 mg thiopentone, followed by 100 mg suxamethonium, the trachea was intubated. The patient was ventilated with nitrous oxide in oxygen and 1.5 vol% of halothane for several minutes in the induction room. After being connected to a new circulatory system in the operating theatre, the patient was ventilated but, in addition inspiratory and expiratory anaesthetic gas concentrations were measured. Despite a vaporizer position of 1.5 vol% the inspiratory concentration of halothane was below 0.2 vol%. Disconnecting the tube, the typical odour of halothane was missing in the inspiratory line of the circulatory system, but was present in the fresh gas tube. Furthermore, the lower part of the soda lime canister was surprisingly hot. After removing both the absorbers, the inspiratory halothane concentration immediately normalized. The absorbers were replaced by canisters filled with fresh soda lime, and the anaesthesia was terminated without further complications. An absorption of halothane by dried soda lime was suspected. METHODS OF SIMULATION. In the first simulation four circulatory systems with two soda lime canisters each were perfused with 21 of oxygen for 48 h. In the second simulation four soda lime canisters placed in one circulatory system were perfused with 1 l for 120 h. For measurement of halothane absorption each canister was placed in a circulatory system. The canister was perfused with a fresh gas flow of 2 l of oxygen and a vaporizer position up to 1.3 vol% of halothane. By the time an equilibrium was reached, i.e., in- and outflow concentrations of halothane were equal for a 3-min period, further halothane vaporization was stopped. In a 30-s interval the soda lime temperature and the gas concentration entering and leaving the soda lime canister were registered. Subsequently, the humidity of the soda lime was determined. RESULTS. In the first simulation 6 of the 8 canisters showed a humidity of soda lime of 15.5% of 19%, with halothane being absorbed in one case. Normally, the equilibrium between in- and outflow gas concentration was reached after 3.5-4 min. In the remaining two canisters the humidity was 14% and 9%. Only a small amount of halothane vapour was absorbed. The halothane concentrations were in equilibrium after 10 to 13 min. The probes from the second simulation revealed a humidity of soda lime of 2% to 8.7%. Below a humidity of 4% the concentration of halothane leaving the canister was greatly reduced. It took 63 min to reach a steady state in the 2% humidity probe and the temperature rose to 43.1 degrees C. CONCLUSIONS. Accidental perfusion of the circulatory system with dry oxygen can cause a reduction in the humidity of soda lime. Dried soda lime delays the increase of halothane concentration in the inspiratory limb. The absorption of halothane is accompanied by an increase in the temperature of the soda lime. Therefore, in every situation lacking a sufficient anaesthetic level during inhalation anaesthesia, absorption of the vaporized anaesthetic must be excluded. Only dried soda lime can absorb halothane.