Abstract: The genus Isopyrum was established by Linné in 1742.  In Species Plantarum ed. I: 557 (1753) he discriminated three species under this genus, i.e., Isopyrum fumarioides, I. thalictroides and I. aquilegioides.  The first was, however, separated by Reichenbach as a monotypic genus-Leptopyrum in 1828, and according to Drummond & Hutchinson the third species I. aquilegioides proved to be Aquilegia viscosa Gouan.  Thus, the only remaining species, I. thalictroides Linn. naturally constitutes the type of the genus Iso- pyrum.   Since then, many new species had been discovered, some of which were later se- parated as distinct genera by different authors.  In  1920  Drummond  &  Hutchinson published an elaborated revision of the genus Isopyrum, where, when recognizing Lepto- pyrum, Souliea, Semiaquilegia as distinct genera,  and also  restoring  the  validity  of Enemion, they furthermore proposed two new genera:  Asteropyrum and Paraquilegia.   In addition to the seven genera above mentioned, Ulbrich further established a new genus Paropyrum in  1925, and another new genus Urophysa in 1929, basing on Iso- pyrum anemonoides Kar. et Kir. (=Paraquiligia uniflora [Aitch.]  Drumm. et Hutch.) and I. henryi Oliv. (=Semiaquilegia henryi [Oliv.] Drumm. et Hutch.) respectively.   Our present study has shown that Asteropyrum, Paraquilegia,  Leptopyrum,  Semia- quilegia, Souliea, Enemion, Usophysa are all remarkably distinct genera.  As for genus Urophysa, it reveals a very close affinity to Aquilegia in having stamanoides inside the fertile stamens, the long-styled carpels and the similar  nectariferous  petals;  but  the more simply divided leaves and the less conspicuous nectariferous organs seem to show that the genus is more primitive than Aquilegia. After a careful examination of the species of genus Isopyrum Linn. s.1. and of its near allies, we find that I. thalictroides Linn., I. anemonoides Kar. et Kit. (=Paropyrum anemonoides [Kar. et Kir.] Ulbr.) and I. manshuricum Kom. (=Semiaquilegia manshurica Kom. and I. manshuricum Kom.) are homogenous both in habit  and  floral  structure, thus forming a very natural group.  The genus Isopyrum Linn. itself, while remaining sixteen species, however, form another natural group, which is easily distinguished from Isopyrum by a series of important characteristics, for which we propose a  new  genus, Dichocarpum.  On account of the foregoing reasons,  it  is  suggested  that  Paropyrum Ulbr. will not uphold as a separate genus, and we also feel necessary to amend the cir-cumscription of Isopyrum as construed by Drummond a