In 12 uremic patients, simultaneous records of the ballistocardiogram and the carotid pulse derivative were taken before and after 36 dialyses. The ECG was also taken in many cases. In the great majority of instances cardiac function remained normal or was unchanged by the dialyses. In a few cases it was definitely improved. However, in 8 instances the heart became weak and (or) incoordinate after dialysis, and in a few patients the abnormality was very striking. These adverse changes in cardiac function took place despite improvement in the blood levels of potassium and other electrolytes. Fortunately, our cases soon recovered from the cardiac damage. We must emphasize the interesting negative aspects of our results. The changes in plasma potassium encountered in our studies are among the largest to be found in the clinic, but they were not accompanied by consistent changes in the circulation. Repeatedly, great improvement in plasma potassium levels was not accompanied by any dynamic change in cardiac function detectable by our methods. Nor did the ECG change consistently. So the expectations generated by classic physiological experiments that abnormal levels of potassium have a deleterious effect on the circulation has no support from our results, probably because the situation of our patients, who had suffered from abnormal blood levels of electrolytes for months or years, was so very different from that in the acute experiments of the physiologists.