Two sensitive procedures to quantitate human exposure to endogenous N-nitroso compounds (NOC) and/or methylating agents have been developed. One, the NPRO test, is based on the excretion of N-nitrosoproline (NPRO) and other N-nitrosoamino acids in the urine, that are measured as an index of endogenous nitrosation, following ingestion of precursors. The NPRO test has been applied to human subjects in clinical and epidemiological studies, and the kinetics and dietary modifiers of endogenous nitrosation have been investigated. Results obtained after application of the NPRO test to subjects at high risk for cancers of the stomach, esophagus, oral cavity and urinary bladder are summarized. In most instances, higher exposures to endogenous NOC were found in high-risk subjects, but individual exposure was greatly affected by dietary modifiers or disease state. Vitamin C efficiently lowered the body burden of intragastrically formed NOC. In experimental animals 3-methyladenine (3-MeAde) is excreted in urine following exposure to methylating NOC. Humans normally excrete 3-MeAde, the origin of which remains unknown. Recently developed analytical methodology permits large numbers of human urine samples to be analyzed and a wide variation is observed. Preliminary results suggest a weak correlation between basal NPRO excretion and background 3-MeAde excretion. Taken together, the results point to an etiological role of endogenously formed NOC in certain human cancers, and provide an interpretation of epidemiological findings that have shown protective effects of fruits and vegetables against several malignancies.