AbstractIt was generally assumed that the ventricle of the primitive vertebrate heart was composed of trabeculated, or spongy, myocardium, supplied by oxygen‐poor luminal blood. In addition, it was presumed that the mixed ventricular myocardium, consisting of a compacta and a spongiosa, and its supply through coronary arteries appeared several times throughout fish evolution. Recent work has suggested, however, that a fully vascularized, mixed myocardium may be the primitive condition in gnathostomes. The present study of the heart ventricles of four holocephalan species aimed to clarify this controversy. Our observations showed that the ventricular myocardium ofChimaera monstrosaandHarriotta raleighanaconsists of a very thin compacta overlying a widespread spongiosa. The ventricle ofHydrolagus affinisis composed exclusively of trabeculated myocardium. In these three species there is a well‐developed coronary artery system. The main coronary artery trunks run along the outflow tract, giving off subepicardial ventricular arteries. The trabeculae of the spongiosa are irrigated by branches of the subepicardial arteries and by penetrating arterial vessels arising directly from the main coronary trunks at the level of the conoventricular junction. The ventricle ofRhinochimaera atlanticahas only spongy myocardium supplied by luminal blood. Small coronary arterial vessels are present in the subepicardium, but they do not enter the myocardial trabeculae. The present findings show for the first time that in a wild living vertebrate species, specificallyH. affinis, an extensive coronary artery system supplying the whole cardiac ventricle exists in the absence of a well‐developed compact ventricular myocardium. This is consistent with the notion derived from experimental work that myocardial cell proliferation and coronary vascular growth rely on distinct developmental programs. Our observations, together with data in the literature on elasmobranchs, support the view that the mixed ventricular myocardium is primitive for chondrichthyans. The reduction or even lack of compacta in holocephali has to be regarded as a derived anatomical trait. Our findings also fit in with the view that the mixed myocardium was the primitive condition in gnathostomes, and that the absence of compact ventricular myocardium in different actinopterygian groups is the result of a repeated loss of such type of cardiac muscle during fish evolution.