1. Potent antisera and test antigens (giving maximum titers of 1:25,600) were prepared from 34 species and strains of fungi by using fractions soluble in 0.85% NaCl solution. 2. The fractions of some organisms exhibited sufficient specificity to permit differentiation, but in most cases cross precipitin reactions were so strong that identification was impossible. In a few cases absorption of precipitins differentiated fungi not separable by the precipitin test. 3. Thus in a limited number of reciprocal tests, members of the Pezizales (Sclerotinia species) were differentiated from members of the Hypocreales (Neurospora tetrasperma, Fusarium species, and a species each of Cylindrocarpon and Ramularia), certain ones of which were in turn distinguished from one another. Strong group reactions occurred among all members of the genus Fusarium tested. The precipitin test did not permit as sharp differentiation between the monilioid strains of N. tetrasperma and the Fusarium group as it did between Fusarium species and the conidial stages of Cylindrocarpon album and a Ramularia sp. 4. Two genera of the Pezizales (S. sclerotiorum and S. fructicola) and two strains of S. fructicola, as well as the plus and minus strains of N. tetrasperma, were differentiated by means of the precipitin absorption test. 5. Many attempts to demonstrate specific precipitabilities for the saline extracts of certain cultures recognized as strains and species of Fusarium (some of which are recognized as conidial stages of different genera of Ascomycetes) were unsuccessful. 6. At present, therefore, the procedure described in this paper is not invariably usable for identification of every one of the fungi tested, although consideration of all reactions of every organism leads to the conclusion that each is a distinct serological entity.