RNA polymerase I (or A) was extracted from nuclear, nucleolar and nucleoplasmic fractions, and resolved into IA and IB forms on a phosphocellulose column. During the course of cycloheximide treatment, the activity of RNA polymerase IB decreased in the nucleoli with concomitant increase in the nucleoplasmic fraction, suggesting strongly that cycloheximide induced specific leakage of IB enzyme from the nucleolus. The activity of IA enzyme did not change in the nucleoli. When nucleoli were incubated in the presence of actinomycin D, all the IA enzyme activity and approximately 30% of the total IB enzyme activity were released in the incubation medium, whereaa 70% of IB activity remained associated with the nucleolar pellet where no IA activity was detected. The enzyme which was released into the incubation medium was tentatively designated as free or unbound RNA polymerase I and that which was associated with the nucleolar pellet as template-bound enzyme. During the treatment with cycloheximide, the activity of bound enzyme, which contained exclusively IB form, decreased rapidly, with kinetics almost identical to that of nucleolar RNA synthesis in vivo. The activity of free enzyme did not change appreciably. At 2 h after partial hepatectomy, IB enzyme activity in the free RNA polymerase fraction increased to almost twice the control, while the bound enzyme activity did not increase appreciably until 4h of regeneration. Enhancement of nucleolar RNA synthesis in vivo was not apparent at 2 h but became significant by 4 h after partial hepatectomy. These results strongly suggest that (a) the above-mentioned procedure is actually fractionating RNA polymerase I into free and bound forms, (b) RNA polymerase IB is the transcriptionally active form in vivo, (c) RNA polymerase IB exists in excess in the nucleoli, but the amount of bound IB molecules, which are engaged in transcription in vivo, must be determined by some other factor(s) than the mere concentration of IB enzyme in the nucleolus, and (d) IA form is not an artifact of isolation but is always present in vivo at a certain amount, although the exact nature of this molecule is not known at present.