The mutagenic activity of the root canal sealing cement, AHPlus, was tested in a bacterial gene mutation assay (Ames test). The material was mixed according to the manufacturer's instruction and tested immediately after mixing and after a setting time of 24 h at 37 degrees C in a humidified chamber. The set material was powdered and both the freshly mixed and the powdered material were eluted in dimethyl sulfoxide (DMSO) and physiological saline (0.1 g/2 ml) for 24 h at 37 degrees C. Aliquots of serially diluted eluates were then used in the standard plate incorporation assay. The Salmonella typhimurium tester strains TA98 and TA100 were employed to detect the induction of frameshift mutations and base pair substitutions both in the presence and in the absence of a metabolically active microsomal fraction from rat liver (S9 fraction). No mutagenic and no toxic effects were found with physiological saline eluates of the freshly mixed material and of mixed material which was set for 24 h. However, DMSO eluates of the freshly mixed AHPlus were mutagenic in tester strain TA100 in a dose-related manner in the absence of S9. A four- to fivefold increase of the mutation frequencies was induced by 2.5 mg AHPlus per plate compared with the number of spontaneous mutants. The mutagenic effect was completely abolished in the presence of a metabolically active S9 fraction. Also, no mutagenic effects were observed with DMSO eluates of AHPlus set for 24 h. However, the set material was more toxic towards bacteria than the freshly mixed sealer. This difference was indicated by a tenfold lower amount of material necessary to cause complete absence of the background lawn in both S. typhimurium tester strains. Therefore, we conclude that at least two different compounds of AHPlus are biologically active in DMSO eluates to cause mutagenic and toxic effects in S. typhimurium TA100 and TA98.