These studies were designed to determine if RIDP was present in a particulate fraction of brains from patients with ALS and PD. Evidence that we have detected RIDP is as follows: (a) DNA polymerase activity persists in the presence of concentrations of actinomycin D and distamycin that inhibit most DNA-directed DNA synthesis (25); (b) the majority of endogenous DNA polymerase activity is sensitive to prior treatment with RNase; (c) the early reaction product is a 4-5 S DNA heteropolymer joined by hydrogen bonds to an RNA molecule; and (d) the purified [3H]DNA product anneals to RNA extracted from the enzyme-containing pellet more extensively than to normal brain RNA or poly(rA). The enzyme activity is in a cytoplasmic particle that can be sedimented at high speed and has the buoyant density of RNA tumor viruses (1.16-1.18 gm/ml). This particulate fraction is not disrupted by physical manipulation and maintains its characteristic density with repeated centrifugations. Treatment with the nonionic surfactant Sterox changes the buoyant density of the enzyme-containing particle to 1.24 gm/ml, the density of the onconavirus virion core. Synthesis of RNA-DNA hybrids by an endogenous reverse transcriptase reaction was found only in normal and diseased Chamorro brains. Examination of a limited number of normal and diseased brains from individuals who lived in the United States produced negative results (39). Definitive characterization of this polymerase activity and identification as a true viral polymerase will depend on purification of biochemically active quantities of this polymerase to determine its template specificities, its cation preference, the fidelity of its transcription product, as well as its antigenic relationship to animal virus and human leukemic RIDP. Of critical importance in these studies will be differentiation of this activity from normal brain DNA polymerase gamma and terminal deoxynucleotidyltransferase.