Some ways in which four short-term tests may be used to evaluate the mutagenicity of drinking water were explored by testing raw and treated water from Lake Bloomington, which serves the town of Bloomington, Illinois (population, 44,000). The water was collected from February 1976 to October 1977 and was concentrated by evaporation or by use of XAD-2 resin. The water was tested for the ability to induce reverse mutation in a prokaryote, Salmonella typhimurium; forward mutation in a mold, Neurospora crassa; mitotic gene conversion in a yeast, Saccharomyces cerevisiae; and reverse mutation in maize, Zea mays. Because of the large number of water samples (54) and the limited amounts of the samples, it was not possible to test all samples in all four tests by all the protocols. Thus, the sensitivities of the four tests to potential mutagens in the water samples could not be rigorously compared. However, the results do show that lake and tap water samples collected during 1976 were toxic but not mutagenic in N. crassa and neither toxic nor genotoxic in S. cerevisiae; lake water collected during 1977 was mutagenic in one line of Z. mays and slightly mutagenic in S. typhimurium strain TA1536 in the presence of rat liver S9. The results suggest that tests that detect a variety of genetic end points should be used when testing complex mixtures such as drinking water. The advantages and disadvantages of the tests and protocols are discussed in terms of their applicability to the study of the mutagenicity of drinking water.