The toxic manifestations of modern intravascular contrast agents have been extensively studied in many parts of the body (1–4). In most areas their predominant toxic effect seems related to their hypertonicity (1, 5, 6). The purpose of this paper is to report the toxic effects of the modern intravascular contrast agents on the sinus node of the canine heart. Methods And Materials An experimental model was designed for this study by modifying one used by Dr. Thomas James for the study of pharmacologic agents (7). The heart was exposed through an anterior bilateral thoracotomy in the fourth intercostal space while the respiration was supported with an intratracheal tube and a respirator. The pericardium was opened over the right coronary atrioventricular groove, and the sinus node artery was identified arising from the right coronary ar-tery, coursing dorsally and then cephalad to perforate the sinus node tissue. The sinus node artery was then ligated just distal to its origin, and a small cannula (I.D. 0.015) was tied in place beyond the ligature (Fig. 1). The injections were made by means of a pair of Harvard infusion pumps connected through a three-way stopcock to the cannula in the sinus node artery. One of the pumps was used for a timed infusion at a known rate of the substance to be tested. The other was used for a timed flushing of the tested substance with normal saline or Ringer's solution. The following data were recorded on a polygraph: one lead of the electrocardio-gram; cardiac rate; aortic pressure; and time of injection or flush. This provided a model for selective perfusion of the sinus node tissue with contrast agents. The right coronary artery remained intact, and the electrocardiogram was not altered by this preparation of the sinus node artery. Between injections into the sinus node artery the collaterals from the atrial branches and from the circumflex branch of the left coronary artery supplied blood to the perfused area as indicated by back bleeding into the cannula. Studies were performed in 45 dogs. Injections of the test substances were made in ordered and randomized series in varying concentrations. Intermittently, saline was used in fixed concentrations as a control to validate the continued usefulness of the particular animal. The return of the animal's circulation to a stable rate was used as a criterion for beginning the next injection. At least two minutes were also allowed between injections. Subsequently, the experiment was extended by atropinizing the animals (1 mg/kg) and/or cutting the vagus nerves after the typical response had been obtained with a test substance injection. These animals were then reinjected with the same test substance. Results 1. Type of Infusions Eliciting a Response: Two types of responses were observed, one to variations in speed of injection, the other to variations in the composition of the injectate.