10?m cytochalasin D (CD) was used to treat cardiocytes from adult human atrium and adult rat atrium and ventricle in long-term primary cultures. Durations of treatment were 0.5, 1, 6, 12, 24 and 48h. In some cultures, the medium containing CD was removed at the planned time to be replaced with the medium without CD.These dishes were then cultured for an additional 48h. Control cultures were exposed to dimethyl sulfoxide (DMSO), the vehicle used to dissolve CD. All cultured cells were first stained with rhodamine-labelled phalloidin to show F-actin and then stained with fluoreseein-labelled antitubulin to show microtubles. The freshly isolated and rounded cardiocytes did not respond to CD, while the spreading cells responded apparently. The actin filament bundles in the peripheral zone of spreading cells were cut into segments. Most segments were gathered into aggregates and granules. Some aggregates were lodged on the inner aspect of the sarcolemma. Small vacuoles were seen between the myofibrils or somewhere along the course of the myofibrils. The CD response from the atrial spreading cells, especially the human atrial spreading cells, were more obvious than that from the rat ventricular cells. Cells exposed to CD and then cultured in normal medium for 48h did not return to normal. Microtubules were not directly affected by CD, but in places where vacuolization occured they made way for vacuoles. All the control cultures made no response to DMSO. The above-mentioned results suggest that different sensitivity to CD existed in cultured adult cardiocytes between different species and that a difference also existed in the contractile machinery between the atrial culturing cells and ventricular cultureing cells of the same species.