Light microscopic studies of the pollen of 51 genera and 151 species of Flacourtiaceae were made from KOH-acetolyzed specimens. The pollen is always found as single grains and is usually triaperturate with compound apertures, isopolar, and psilate to reticulate. Within the triaperturate type, the pollen of the family ranges from unspecialized to moderately specialized. Differences in pollen morphology do not often serve to delineate the taxonomic tribes. The Berberidopsideae and Oncobeae show a considerable range of size and sculpturing patterns. Scolopieae, Banareae, Homalieae, and Flacourtieae have pollen which can scarcely be distinguished. Pollen of Casearieae is similar to that of the preceding group of tribes but adds more specialized trends. Both Pangieae and Paropsieae are isolated in the family, and the Paropsieae should be assigned to the Passifloraceae. Comparisons with families of putatively related orders indicate that the relationships of the Flacourtiaceae based on pollen morphology accord well with Takhtajan's system.By one recent estimate ( Hutchinson, 1967 ) , the Flacourtiaceae include about 90 genera and 1284 species. This large, tropical family has received little attention by taxonomists and morphologists in proportion to its size and evolutionary importance. Such neglect seems to be due to the complex taxonomic history of the family and its reputation as a taxon in which to place many genera of uncertain status (cf. Williams, 1965). As understood by Warburg (1893) and Gilg (1925) the family is quite heterogeneous and indeterminate. The concept of the family has been greatly improved by the generic treatment of Hutchinson ( 1967 ) .The Flacourtiaceae include some very primitive genera in the tribes Berberidopsideae and Oncobeae (Hutchinson, 1967), and the family usually occupies a basal position in the Violales (Cronquist, 1968;Takhtajan, 1969), Cistales (Gundersen, 1950; Thome, 1968) or Bixales (Hutchinson, 1959). In most of these schemes the Violales are considered close to the Theales, but their relationship to Ranalian lineage is still controversial (Takhtajan, 1969;Meeuse, 1970). A brief resume of suggested affinities has been presented by Angely (1966). As many as nine orders have been suggested to be derived from flacourtiaceous tock (Takhtajan, 1969), and many of these suggestions will probably prove correct. However, in spite of the potential importance of microscopic characters in interpreting evolutionary trends, such observations are so scarce in many families as to make many statements of taxonomic affinity with Flacourtiaceae fall within the realm of speculation. Recognizing that pollen provides only one line of evidence in phylogenetic studies (Thorne, 1963), I have begun with a survey of the light microscopic features of the pollen in order to provide an s X I am grateful to the curators of the U. S. National Herbarium and of the Missouri Botanical Garden for permission to use the collections and for their assistance. The SEM photomicrographs were made possible through the ...