It is possible in many situations to identify humans exposed to potentially toxic materials in the workplace and in the environment. As in most human studies, there tends to be a high degree of interindividual variability in response to chemical insults. Some non-exposed control individuals exhibit as high a level of damage as some exposed individuals and some of these have levels of damage as low as many of the controls. Thus, it is only the mean values of the groups that can substantiate an exposure-related problem; the data on an individual basis are still of limited use. While human lymphocytes remain the most popular cell type for monitoring purposes, sperm, buccal, nasal, epithelial and placental cells are also used. However, for interpretation of responses, the issue of confounding factors must be addressed. There are endogenous confounding factors, such as age, gender, and genetic make-up and exogenous ones, including lifestyle habits (smoking, drinking, etc.) There are biomarkers of exposure, effect/response and susceptibility and the last may be influenced by the genotype and polymorphism genes existing in a population. From our own studies, confounding effects on cytogenetic damage and ras oncoproteins will be considered in relation to workers exposed to vinyl chloride and petroleum emissions and to volunteers taking Vitamin C supplementation. Smoking history, exposure and duration of employment affected the worker studies. For petroleum emissions, so did gender and season of exposure. For the non-smoking volunteer Vitamin C supplementation study, cholesterol levels, plasma Vitamin C levels, lipid peroxidation products and DNA damage in the Comet assay were also measured. Gender affected differences in Vitamin C levels, antioxidant capacity and the number of chromosome aberrations induced by bleomycin challenge in vitro. The results were the same for both high and low cholesterol subjects. The relationship between biomarkers and the various factors which affect them is complex. Sometimes the variables are not completely independent of each other.