A number of aprotic solvents which had previously been found to induce mitotic aneuploidy in yeast were tested for their effects on re-assembly of twice recycled tubulin from pig brain. Some of the solvents which were strong aneuploidy-inducing mutagens in yeast slowed down tubulin assembly in vitro at concentrations lower than those required for aneuploidy induction. Ethyl acetate, methyl acetate, diethyl ketone and acetonitrile fell into this category. Other strong aneuploidy-inducing agents like acetone and 2-methoxyethyl acetate accelerated tubulin assembly. Non-genetically active methyl isopropyl ketone and isopropyl acetate both accelerated assembly, whereas methyl n-propyl ketone and n-propyl acetate were weak inducers of aneuploidy and slowed down the rate and extent of assembly. Those chemicals which slowed down the assembly rate also reduced the extent of assembly. Most chemicals which accelerated assembly also led to an increased extent of assembly, with the exception of isopropyl acetate. At the higher concentrations, however, a maximum assembly rate was reached which was followed by a slow decline. Although a perfect correlation between effects on the induction of chromosomal malsegregation and the interference with tubulin assembly in vitro was not seen, the experiments with tubulin were carried out using this class of chemicals because some of them strongly induced mitotic aneuploidy under conditions which suggested tubulin to be the prime target. The lack of a perfect coincidence might be due to species differences between the porcine brain and the yeast spindle tubulin, or the test for aneuploidy induction may have been negative because the concentrations required for an effect on yeast tubulin may be greater than the general lethal toxicity limit. Bearing this reservation in mind, the results suggest that the yeast aneuploidy test has a considerable predictive value for mammalian mutagenicity.