Various alkali (e.g., soda lime) convert sevoflurane to CF2=C(CF3)OCH2F, a vinyl ether called "Compound A, " whose toxicity raises concerns regarding the safe administration of sevoflurane via rebreathing circuits. In the present investigation, we measured the sevoflurane degradation and output of Compound A caused by standard (13% water) Baralyme brand absorbent and standard (15% water) soda lime, and Baralyme and soda lime having various water contents (including no water). We used a flow-through system, applying a gas flow rate relative to absorbent volume that roughly equaled the rate/volume found in clinical practice. Both absorbents, at similar water contents, temperatures, and sevoflurane concentrations, produced roughly equal concentrations of Compound A. Dry and nearly dry absorbents produced less Compound A early in exposure to sevoflurane, and more later, than standard absorbents. Increases in temperature and sevoflurane concentration increased output of Compound A. Both absorbents, especially when dry, also destroyed Compound A, the concentration exiting from absorbent resulting from a complex sum of production and destruction. We conclude that the variability of concentrations of Compound A found in clinical practice may be largely explained by the inflow rate used (i.e., by rebreathing), sevoflurane concentration, and absorbent temperature and dryness. The effect of dryness is complex, with fresh dry absorbent destroying Compound A as it is made, and with dry absorbent that has been exposed to sevoflurane for a period of time providing a sometimes unusually high output of Compound A.