Although chemical disinfection of drinking water is a highly protective public health practice, the disinfection process is known to produce toxic contaminants. Epidemiological studies associate chlorinated drinking water with quantitatively increased risks of rectal, kidney, and bladder cancer. One study found a significant exposure-response association between water mutagenicity and relative risk for bladder and kidney cancer. A number of studies found that several types of disinfection processes increase the level of mutagens detected by the Salmonella assay. As part of a comprehensive study to examine chlorinated and ozonated/postchlorinated drinking water for toxicological contaminants, the Salmonella mutagenicity assay was used to screen both volatile and nonvolatile organic components. The assay also compared the use of reverse osmosis and XAD resin procedures for concentrating the nonvolatile components. Companion papers provide the results from other toxicological assays and chemical analysis of the drinking water samples. The volatile components of the ozonated/postchlorinated and chlorinated water samples and a trihalomethane mixture were mutagenic to a Salmonella tester strain transfected with a rat theta-class glutathione S-transferase and predominantly nonmutagenic in the control strain. In this study, the nonvolatile XAD concentrate of the untreated water possessed a low level of mutagenic activity. However, compared to the levels of mutagenicity in the finished water XAD concentrates, the contribution from the settled source water was minimal. The mutagenicity seen in the reverse osmosis concentrates was < 50% of that seen in the XAD concentrates. Overall, mutagenic responses were similar to those observed in other North American studies and provide evidence that the pilot plant produced disinfection by-products similar to that seen in other studies.