Coronary development in birds and mammals occurs in concert with myocardial compaction, likely in response to myocardial hypoxia. Furthermore, the degree of compaction of a cardiac chamber greatly reflects its work rate. These same driving forces likely featured prominently during the evolution of the coronary circulation among chordates. Yet, the means of supplying oxygen to fish hearts represent solutions that are far more diverse, possibly more complex and certainly more mysterious than those for the adult mammalian heart. To date, a coronary circulation has always been found associated with compact myocardium in fish; this is true for ventricle and the conus arteriosus. However, most fish species likely do not have a coronary circulation, nor do they have a thickened compact myocardium, and instead rely on the other oxygen supply route for the fish heart, the luminal oxygen supply to spongy myocardium. The archetype for the chambered vertebrate heart was likely avascular because no cyclostome has a coronary circulation. Nevertheless, the coronary circulation likely appeared when the first jawed vertebrates evolved because all extant elasmobranchs possess a coronary circulation that supplies the spongy and compact myocardial layers of the ventricle, as well as the compact myocardium of the conus. Extant species of basal teleosts provide evidence of a progressive evolutionary transition toward a loss of conal myocardium and the development of three forms of ventricular anatomy seen among modern day teleost species. The most prominent form is a reversion back to the archetypal spongy ventricle that lacks a coronary circulation. Most of the remaining teleosts have limited the coronary circulation to the outer compact myocardium and left the spongy myocardium avascular. A few species have a highly developed coronary system that serves both the spongy and compact myocardium, as in elasmobranchs. Thus, beyond the highly developed coronary circulations of endothermic sharks and tunas, cardiac evolution among fishes appears to have moved toward independence from a coronary circulation, beginning perhaps in the cyprinid lineage. Indeed, fish hearts comprise at least 30% and most often 100% spongy myocardium. Although air breathing in fishes increased the security of the luminal oxygen supply to the heart, it did not supplant the need for a coronary circulation. Many mysteries still remain regarding the coronary circulation in fishes including the extent to which the spongy myocardium is vascularized.