Individuals may be exposed to solvent mixtures and fuel either at work or home, through air, water and food contamination. Few studies have addressed the genotoxic effects of mixed, low-level exposure to fuel and solvent. This was an optimally designed study where each subject was sampled prior to exposure and after 15 and 30 weeks while exposed, in a repeated measures design with each subject serving as his own control. Fifty men aged between 18 and 50, working on aircraft equipment operation and maintenance at a military installation were included. Eight unexposed men were concurrently sampled. Sister-chromatid exchanges (SCE) and micronuclei (MN) frequency were measured in conjunction with air sampling and expired breath analysis for jet fuel (JP-4), 1,1,1-trichloroethane, methyl ethyl ketone, xylenes, toluene and methylene chloride. Exposure levels measured by industrial hygiene were very low (all means <6 p.p.m.), <10% of the OSHA standard. Expired breath levels were also low, <25 p.p.b. A small but statistically significant increase in the frequency of SCE occurred after 30 weeks of exposure for sheet metal workers (P = 0.003) and for painters (P = 0.05). The MN frequency in the sheet metal workers initially showed a statistically significant increase, but by 30 weeks had decreased. Cigarette smoking, alcohol and caffeine use were not associated with changes from baseline for either MN or SCE. Smokers, however, had significantly higher values of SCEs at baseline than did nonsmokers. In summary, these findings suggest that small increases in SCEs in particular, may serve as a sensitive biologic indicator of low level hydrocarbon exposure in as much as statistically significant changes occurred in the highest exposed groups but not in the low or no exposure groups. Chance occurrence or exposures to other occupational or non-occupational agents cannot be eliminated as a cause of the study findings.