This study investigates the possible role of impurities in dibromochloropropane in inducing mutations, and discusses the importance of contaminants in mutagenicity and carcinogenicity testing. A technical grade sample and a pure sample of DBCP (no epichlorohydrin added) were assayed in Salmonella typhimurium TA1535, with and without S-9 activation, using agar overlay procedures and dessicator procedures. Assays performed with both technical and pure DBCP without metabolic activation resulted respectively in an increase in revertants with increasing dose (0-1600 microgram/plate) when the technical grade was tested, and no substantial increase in revertants over the same dose range when the pure DBCP was tested. Distillation of technical grade DBCP yielded an initial fraction containing high amounts of epichlorohydrin (verified by GC-MS) which was highly mutagenic. The amount of epichlorohydrin in the technical DBCP sample was calculated for each dose level tested, and the number of revertants obtained in tests of the technical DBCP sample could be attributed solely to the calculated amount of epichlorohydrin in each test dose. Tests of pure and technical DBCP using a dessicator technique produced a similar differential between the technical and pure compounds in mutagenicity. Activation of both technical and pure DBCP with S-9 from Aroclor pre-treated rats resulted in high mutagenic responses, of equal magnitude, from both preparations.