Cancer patients and tumor-bearing animals excrete high levels of modified purines and pyrimidines some of which, e.g., N2,N2-dimethylguanosine, can originate only from transfer RNA (tRNA). Until recently, it could not be ascertained whether the high level of excretion of such compounds is due to cell death or specific tRNA turnover. However, an approach to this problem became feasible, with beta-aminoisobutyric acid as a probe. This compound is a terminal degradation product of thymine which is present in both DNA and tRNA. Since the pathway of synthesis of thymine is different in the two macromolecules, it and its end product, beta-aminoisobutyric acid can be differentially labeled with [14C]formate and [3H3]methylmethionine as precursors. Therefore the ratio of the two labels in the excreted beta-aminoisobutyric acid is a measure of the macromolecular origin of the degradation product. We have found from such analysis that tRNA's are not homogeneous in their turnover rate. There is a subpopulation that turns over much faster than the rest. The turnover rate of a subpopulation of tRNA's in tumor tissue exceeds the turnover rate of tRNA's in normal tissue. Such rapid degradation of tRNA's must be the source of the massive excretion of modified nucleosides by cancer patients which can be 10-fold higher than in normal subjects.