A mouse monoclonal antibody (Mab) was prepared which showed high specificity for a potential marker of exposure to methylating carcinogens such as 3-methyladenine (3-MeAde). In a low-temperature (4 degrees C) ELISA a linear calibration curve was obtained between 3 and 50 fmol/well. In combination with an immunoaffinity (IA) column prepared from a 3-MeAde rabbit antiserum, the ELISA was used to determine 3-MeAde in urine. The IA-ELISA method was validated by comparison with results obtained by an IA-GC-MS method. The effect of consuming a low 3-MeAde diet on urinary 3-MeAde excretion was investigated in a human volunteer. Urine collected during a 'normal' diet exhibited the characteristic variation in 3-MeAde levels previously observed (9.5 micrograms/24 h, SD = 4.4, n = 5). In contrast, 3-MeAde excretion was markedly lower and less variable on days when the diet was closely controlled (0.63 microgram/24 h, SD = 0.08, n = 3). Dietary intake of 3-MeAde on the latter days was between 0.37 and 0.43 microgram/day, indicating that most, if not all, of the 3-MeAde seen in previous experiments was derived from the diet. The origin of dietary 3-MeAde is not known, but may be related to fumigant use. Dietary manipulation affords the possibility of carrying out model studies, in volunteers, on 3-MeAde intake and formation in vivo.