In the present study pollen morphology of the Euphorbeaceae is treated as an additional character in taxonomy. Besides the greater part of the genera occurring in the system of PAX and K. HOFFMANN (1931), most of the genera published after 1931 are studied. The pollen grains have been described with the aid of a terminology as simple as possible. In principle the terminology of IVERSEN and TROELS-SMITH has been followed, although in addition, many improvements of ERDTMAN have been used. One of the simplifications is the rejection of POTONIÉ’s term sculpture. All elements occurring on the endexine are called structure elements; all structure elements together form the structure of a pollen grain. For the sake of consequence endexine apertures and extexine apertures are discussed separately. Different pollen grains are placed in different pollen types. If the differences are of minor importance, the pollen grains are placed in subtypes. Several types can have some characters in common. To express the correspondences, these types are assembled in configurations. As the pollen types in Phyllanthoideae and Crotonoideae differ distinctly, the division of the Euphorbiaceae in these subfamilies is maintained in the discussion of the results. The Phyllanthodieae can be separated in three large groups of pollen types ( Antidesma configuration, Amanoa configuration and Aristogeitonia configuration), which agrees with the grouping of PAX in 1924. The remaining small configurations belong in taxonomic respect to the genera of the Antidesma configuration. In the Crotonoideae many genera possess pollen grains with a croton-pattern. These genera should be treated as a single group. Besides this natural group, the Plukenetiinae possess pollen grains which are clearly distinguished from other genera in the Crotonoideae. Pollen grains of Omphalea are similar to those in the Plukenetia configuration. This pollen-morphological result agrees with the opinion of CROIZAT. The remaining pollen grains in the Crotonoideae are less easy to differentiate in groups. One of the largest configurations is the Mallotus configuration, which includes most genera of the Acalypheae and several genera or other tribes. The Hippomane configuration is another large one. This configuration comprises the tribes Hippomaneae and Euphorbieae. The pollen grains of both tribes are very similar. The genus Pachystroma is pollen-morphologically as well as taxonomically related to the tribe Hippomaneae. Pera, treated as a separate tribe by PAX and K. HOFFMANN, is related by its pollen grains to some genera in the Acalypheae. Dalechampia is habitually related to the genera in the Plukenetiinae. Pollenmorphological data, however, do not support this relation. The pollen grains of Dalechampia are not similar to any other pollen type. The morphology of the pollen grains of the Stenolobeae is in agreement with the opinion of PAX, that any separation of these Australian genera is an artificial one.