Destruction of microtubules by an antimitotic drug, e.g. vinblastine, leads to inhibition of axonal transport. This process may be important in the maintenance of synaptic viability or in the mediation of local changes in membrane structure during the formation and modification of synapses. The authors wondered, therefore, whether disruption of microtubules with vinblastine could alter the acquisition and retention of learning behaviour. The mice ddN-F10~F12 strain animals of age between 6 to 10 weeks were used. They were handled and allowed open field activity beforehand to get them used to the conditions of the experiments. The standard deviation in learning experiments based on avoidance training is large if inbred strains are used. Therefore, ddN strain mice, which have a large litter size and high training scores were selected for genetic improvement through avoidance training. The standard deviation of scores for the twelfth generation, that is ddN-F12, was significantly different from that of earlier generations. There were no remarkable differences between males and females. The effects of intraperitoneal administration of vinblastine were tested as the changes of body weight and open field activity performed at one week intervals. There were no marked differences between saline and vinblastine injected mice. Vinblastine studies were conducted by giving each animal an intraperitoneal dose of 50 μg in 0.5 ml of saline. Controls were given the same volumes of saline only. In mice given the injection one week before the first avoidance training session, learning ability in the first training session was much less than in the saline injected group. In the second and third training sessions, learning ability gradually increased to reach a normal rate. The mice injected immediately after the first avoidance training session showed a slight retardation of memory retention in the first training session after the injection, and no difference from saline injected animals in the second training session. The same results were obtained for the consolidated animals. Vinblastine did not cause any loss of memory in the first and second training sessions after the injection. In mice given the injection one week before the first reward training session, learning ability in the first training session was slightly increased, but in the second reward training session, memory acquisition was slightly decreased. Similar results were obtained for mice injected immediately after the first reward training session. After consolidation of memory, vinblastine did not cause any loss of memory. The mice were given a single intraperitoneal dose of (3)H-vinblastine (40μ Ci in 0.5 ml of saline). Animals were killed one hour, one day, or one week after injection and the brains was fixed. Sections were prepared for autoradiographic analysis. One hour after injection, label was present in the ependymal layer of ventricles. Twenty-four hours after injection, the intensity of label had increased in the ependymal and adjacent regions of the ventricles. One week after injection, grains still remained in the vicinity of the ventricles. The findings suggest that microtubules were deeply involved in memory formation and retention.