Abstract: Wide distribution and abundant production of wild ergots had been reported in China and many collections proved valuable because of their relatively high contents of alkaloids. According to informations so far available, 5 species of ergot were recorded, namely, Claviceps purpurea (Fr.) Tul., C. microcephala (Wallr.) Tul., C. paspali Stev. Hall., C. miscanthi Saw., and C. syntherismae Saw. Of the 70 recorded host species which were distributed in 35 genera, only one species (Carex subpediformis (Kuk) Suto et Suzuki) belongs to Cyperaceae; the rest are all Gramineal, with 13 genera including 33 species distributed in the subfamily Poaeoidae, 9 genera including 19 species distributed in Agrostoideae, 13 genera including 17 species in Panicoideae. None were found in the subfamily Oryzoideae.In Poaeoideae, 25 host species of 9 genera belong to Hordeae, a tribe with largest number of host species; the tribe Agrostideae of Agrostoideae comes next, possessing 19 host species of 8 genera. Regardless of different localities (North and Northeast China), the alkaloid contents of ergot collections from Calamagrostis epigeios and Clinelymus dahuricus were relatively high (0.405—0.56% and 0.30—0.33% respectively).On the other hand, the alkaloid content of rye ergots was rather low (0.06—0.065%). Variations of alkaloid contents due to different localities as well as different years appeared inconspicuous. Consequently, genetic capability of ergot races appeared to be the major factor resposible for alkaloid content, while host species may also have considerable influence. C. microcephala is conventionally distinguished from C. purpurea by the size of perithecial stroma. However, it is apparent in our experiments that this is not a valid criterion because of its extreme variability. As a matter of fact, stromatal size varies proportionally with the size of Sclerotium which in turn is dependent on the size of host seeds. When these two organisms were inoculated to the same host species (rye for instance), the size of stromata of the resultant sclerotia showed virtually no conspicuous difference.Therefore, C. microcephala should not be differentiated from C. purpurea.