It has been clear for many years that the evolution of parasitism in the angiosperms has taken place independently in a number of unrelated groups (Kuijt, 1969). In other words, the parasitic organs (haustoria) which both structurally and physiologically join host and parasite are remarkable instances of convergent evolution in these different groups. The main thread which can be followed throughout the many anatomical studies which in the past have been carried out on haustoria is the search for common denominators between various groups; that is, as to how much the haustoria of unrelated taxa have come to resemble each other. When Solms-Laubach (1867–1868) more than a century ago reviewed the structure of angiosperm haustoria it required well over one hundred pages to do so. Needless to say the bulk of our present knowledge has been obtained since that time, and light microscopy continues to unearth significant new information. In a recent review of haustorial anatomy (Kuijt, 1969), several structural parallelisms are noted. The advent of electron microscopy has, since the latter review, resulted in a great deal of additional information which has enabled certain comparisons between groups to be made. Ultrastructural studies have quite naturally concentrated on the most readily available which, of course, are also those which have been most i ntensively studied with the light microscope in the past. The haustoria of several families, such as Rafflesiaceae, Balanophoraceae, Hydnoraceae, and Lennoaceae, have scarcely been studied even with the light microscope. The purpose of the present paper is to give a general review of present knowledge of the ultrastructure of haustoria. To do so we must of necessity ignore most recent purely anatomical work, except where relevant to the topics under discussion. Ultrastructural observations have been made on mistletoes (Arceuthobium and Phthirusa, Loranthaceae), Santalaceae (several genera), Scrophulariaceae(Castilleja), Orobanchaceae (Orobanche) and Convolvulaccae (Cuscuta), in which order the discussion will proceed.