A diploid yeast strain, D81, was constructed heterozygous for seven recessive markers linked on the left arm of chromosome VII to study the localization of induced mitotic crossing over. The mutagens used were carofur also called nifurprazinum (1-(5-nitro-2-furyl)-2-(6-amino-3-pyridazyl)-ethylene hydrochloride), diepoxybutane, ethylmethanesulfonate, nitrous acid and 1-nitrosoimidazolidinone-2. All agents induced high frequencies of mitotic crossing over at doses exerting only a low degree of killing. The distribution of recombinational events was compared for five intervals. The distribution pattern of spontaneous mitotic crossing over was different from all the patterns obtained after mutagenic treatments. Nitrous acid and diepoxybutane induced the same pattern, which was different from the patterns induced by carofur, EMS and 1-nitrosoimidazolidinone-2. The patterns induced by the latter three mutagens were again different amongst each other. Repeat experiments showed that the patterns induced by a given mutagen were reproducible. Tetrad analysis with a representative sample of segregants induced by diepoxybutane and carofur showed that the treatments actually induced mitotic crossing-over. The pattern of meiotic recombinational events was different from those of spontaneous and mutagen induced mitotic recombination. Inducibility of mitotic crossing-over was low at the proximal and distal ends of the chromosome arm and highest in the middle. Each interval showed a different response to those mutagens that differed in their patterns of induced mitotic crossing over. The observed mutagen specific effects are considered as an indication of mutagen specificity. No plausible explanation for mutagen specificity could be given. However, the data presented reveal the same situation as found in induction of chromosome breaks, as reported by other authors. Apparently, mutagen specificity is quite a general phenomenon even for genetic effects in larger intervals of a chromosome.