1. The pharmacological profile and the anatomical localization of a putative dopamine D4 receptor were assessed in sections of rat and human atria and ventricles using combined radioligand binding and autoradiographic techniques with [3H]-spiperone as a ligand. 2. [3H]-Spiperone was bound specifically to sections of human and rat atria and ventricles. The binding was time-, temperature- and concentration-dependent, belonging to a single class of high-affinity sites. In atria, the dissociation constant value (Kd) was 0.45 nM in rats and 0.32 nM in humans, and the maximum density of binding sites (Bmax) was 31.6+/-2.9 fmol mg(-1) tissue in rats and 18.8+/-0.7 fmol mg(-1) tissue in humans. In ventricles, Kd was 0.38 nM in rats and 0.39 nM in humans, and the Bmax was 43.5+/-3.0 fmol mg(-1) tissue in rats and 56.4+/-3.2 fmol mg(-1) tissue in humans. 3. The pharmacological profile of [3H]-spiperone binding to sections of both rat and human atria and ventricles was consistent with the labelling of dopamine D2-like receptors. [3H]-Spiperone binding was more sensitive to displacement by the neuroleptic clozapine in sections of atria than of ventricles, suggesting the expression of a dopamine D4 receptor in atrial tissue. Moreover, preincubation of some sections with a dopamine D4 receptor antibody and subsequent exposure to [3H]-spiperone caused a remarkable decrease of radioligand binding to sections of atria, but only a slight reduction of binding to sections of ventricles. 4. Light microscope autoradiography revealed the accumulation of silver grains over atrial tissue within atrial myocardiocytes. A higher density of silver grains was developed in rat than in human atria. In ventricles, silver grains were accumulated primarily in cluster areas both in rats and in humans. 5. The above findings suggest the expression of a dopamine D4 receptor in rat atria, but not in ventricles. A similar site was identified in human atria. The possible relevance of a dopamine D4 receptor in the heart is discussed.