We describe the development of an immunofluorescent method for the detection of resistance to agents which affect the integrity of the cellular microtubular network. Three pleiotropic resistant MCF-7 human breast carcinoma cell lines mixed with vaginal adenocarcinoma cells were selected in serially increasing drug concentrations, and demonstrated a 30-fold increase in resistance to colchicine. Transport studies indicated that there was no difference in drug accumulation between the sensitive and resistant lines. The colchicine-binding capacity of cell extracts from sensitive and resistant cells was similar (Kd for sensitive cells was 1.9 x 10-6M and for resistant cells 1.58 x 10-6M). There were, however, significant differences in cytoskeletal morphology between sensitive and resistant cells. Drug-sensitive cells were mostly large (about 70 μm2) and flattened. Their cytoplasm was filled with a microtubular network in which, in most of the cases, single fibers could be differentiated. Cells usually had a microtubule-organizing center and paracortical bundles of microtubules. In contrast, drug-resistant cells were mostly rounded and grew in clumps. In only 40% of these cells could single microtubular fibers be differentiated. Resistant cells lacked a microtubule-organizing center and had no clear paracortical bundles of microtubules. The tubulin-binding agents tested caused a sequence of morphological changes in sensitive cells. These changes included precipitation of tubulin and disappearance of cytoskeletal structure. Changes occurred initially within 3 h of incubation, but were expressed in all cells after 6 h. If, after 3 h of drug exposure, cells were subcultured in drug-free media, the cytoskeletal structure reformed within 10 h. Maximal recovery of cytoskeletal structure occurred 22 h after drug removal and was sustained up to 36 h. In contrast to changes observed in sensitive cells, drug exposure did not induce changes in the morphology of cytoskeleton in resistant cells. Cells from all three resistant lines reverted to sensitivity after 7 months of culture in drug-free media. This was first detected by immunofluorescence and then confirmed by cloning assay. Since the cytoskeletal disintegration of sensitive cells is readily detectable within a few hours of in vitro drug treatment, immunofluorescent imaging may have its clinical application in predicting the sensitivity/resistance to microtubule-binding agents