Several investigations have recently been published which suggest that classical tricuspid atresia is produced by an imperforate valve membrane interposed between the right atrium and the underlying hypoplastic right ventricle. Since this is contrary to our understanding of the anatomy of the majority of examples of tricuspid atresia, and since a knowledge of the basic anatomy is vital for correct clinical diagnosis, particularly using two-dimensional echocardiography, we have reviewed the morphology of the lesions which can produce "tricuspid" atresia. For this purpose, we have considered all those lesions which produce blockage of direct exist from the morphologically right atrium in patients with situs solitus. In some instances, the lesion may not in fact affect a morphologically tricuspid valve, and for this reason right atrioventricular valve atresia is a more accurate term for the group. But because of the widespread clinical use and understanding of tricuspid atresia as describing anomalies with no other exit from the morphologically right atrium, we have used this term. In essence, tricuspid atresia as thus defined can be due to either an imperforate atrioventricular valve membrane or to absence of the right atrioventricular connexion. Contrary to widely held opinion, the commonest type of classical tricuspid atresia is due to absent right connexion. In this form, atrioventricular sulcus tissue extends in to the septal atrioventricular junction, completely separating the floor of the right atrium from the ventricular myocardium. The right ventricular chamber is rudimentary, lacking an inlet. Cases with atrioventricular concordance and an imperforate tricuspid valve can exist but are rare. Such cases, when associated with Ebstein's malformations, may have right atrioventricular annuli sufficiently large to permit insertion of a prosthetic valve, so are of major clinical significance. Rarer forms of tricuspid atresia can be produced with either absent connexions or imperforate membranes in association with other chamber combinations. We have sectioned heart specimens to show the salient clinical anatomy of these various forms, and constructed diagrams to show the features of the rarer types. Finally, we have discussed the relationship of these various forms of tricuspid atresia to the univentricular heart, showing how most disagreements relate to semantics.