The reported presence of large quantities of a high molecular weight form of angiotensin I and II in renin granules from rat kidney cortex was investigated. Subcellular fractionation by differential centrifugation and isopycnic gradient centrifugation confirmed the presence of angiotensin I immunoreactive material, but the distribution resembled that of lysosomes rather than renin granules, mitochondria or protein. The material did not possess pressor properties, was stable to incubation at 37.degree. C and was precipitated by ethanol. Incubation of subcellular fractions with peptidase inhibitors known to inhibit the activity of renal angiotensinases (EDTA, diisopropyl phosphorofluoridate and 2,3-dimercaptopropanol) decreased the level of angiotensin I immunoreactive material in kidney fractions treated by radioimmunoassay. By paper chromatography subcellular fractions were apparently capable of degrading 125I-labeled angiotensin I used as tracer in the radioimmunoassay. The degree of degradation followed the distribution of lysosomes among the fractions and was decreased by the angiotensinase inhibitors. The apparent MW of the major portion of angiotensinase activity persisting despite the presence of angiotensinase inhibitors was 75,000, a value similar to that of angiotensin immunoreactive material in kidney fractions treated with non-ionic detergent. On the basis of the present findings, angiotensin immunoreactive material may be largely artifactual, being created by the hydrolysis of tracer by angiotensinase enzymes during radioimmunoassay to form fragments that are no longer capable of binding to the specific antibody. This would produce results that would appear the same as those produced had genuine angiotensin immunoreactivity indeed been present in the samples. Such an effect could, in principle, occur in any competitive protein binding assay and should be tested.