Determination of mutagenic activity in bacterial systems has become accepted as an initial step in the evaluation of the carcinogenic potential of new chemicals. In this paper, a bacterial mutagen screening technique is described in which chemicals can be tested in 10 tester strains over a 10,000-fold concentration gradient both with and without metabolic activation. Using this assay, 855 chemicals were tested, and 182 were found to be mutagenic in one or more of the tester strains. Included were 299 chemicals used in chemical manufacturing or laboratory synthesis. Of these, 20% gave a positive response in one or more strains. The high rate of positives undoubtedly reflects the high chemical reactivity of compounds in this group. In contrast, when 261 organic chemicals which were synthesized for evaluation as potential pharmaceutical or agricultural products were tested, only 8% were identified as mutagenic. The Salmonella typhimurium tester strains TA98 and TA1538 proved to be very reliable and efficient in detecting and identifying frame-shift mutagens. TA100 was the most sensitive tester strain, detecting 142 of the 182 mutagens encountered in the study. However, since TA100 detected both base substitution mutagens and frame-shift mutagens, this tester strain was not suitable for the specific identification of base substitution mutagens. Base substitution mutagens were more reliably detected by Escherichia coli tester strains WP2 and WP2 uvrA- than they were by S. typhimurium strains G46 and TA1535. The data obtained when mutagens are tested by the concentration gradient procedures can include (a) the activity spectrum in tester strains, (b) identification as either frame-shift or base substitution mutagens, (c) the minimal concentration at which auxotroph growth is inhibited, and (d) mutagenic potency in terms of minimal concentration at which mutagenicity is observed. The data obtained have been found to be of immediate use. For example, with manufacturing intermediates the data have been combined with other toxicity data and used as a basis for setting safety standards for handling such compounds in the workplace. In addition, positive bacterial mutagenicity data on selected members of new series of organic compounds can serve to alert the chemist early to the possibility that the compounds may possess undesirable toxic properties, particularly carcinogenicity. Also, this type of data should be of great value both in the planning and in the interpretation of other in vitro tests designed to evaluate the potential carcinogenicity in mammals of chemicals found to be positive in bacterial tests.