Many anticancer agents have been shown to be carcinogenic, mutagenic and teratogenic in experimental animals and in in vitro test systems. Epidemiological data on the association of second neoplasms with a specific chemotherapy treatment is available on some 30 agents, and in the case of 10 compounds the overall evidence on human carcinogenicity has been evaluated to be conclusive (Group 1: IARC, 1987 and 1990). The primary source of human exposure to anticancer drugs is from their use in therapy of cancer. However, persons employed in the manufacture, preparation and administration of the drugs to patients and in nursing patients may also be exposed. Safe handling of anticancer drugs, since the introduction of various general handling guidelines, is now good practice in hospitals, pharmacies and drug manufacturing companies of most developed countries. Careless handling of cancer chemotherapeutic agents may lead to exposure of the personnel in amounts detectable with chemical or biological methods in the body fluids or cell samples of the subjects. The exposure is typically to mixed compounds over long-term and to low exposure levels with accidental peaks. Therefore, the use of biological exposure markers is appropriate for the monitoring of such exposure patterns. The biological markers/methods for exposure assessment are either non-specific (e.g., cytogenetic damage, point mutations or 32P-post-labelling adducts in peripheral blood lymphocytes, urinary mutagenicity) or specific for a given compound (immunological methods for DNA adducts, specific analytical methods). Studies have revealed minor amounts of cyclophosphamide in the urine of pharmacy technicians and nurses handling the drug even when taking special safety precautions (Sessink et al. (1994a) J. Occup. Med., 36, 79; Sessink et al. (1994b) Arch. Env. Health, 49, 165). Another study showed surface wipe samples with measurable cyclophosphamide even away from the handling site (McDevitt et al. (1993) J. Occup. Med., 5, 57). These studies strongly implicate the importance of skin absorption as an exposure route. Also accidental spillage is never completely avoidable (Sorsa et al. (1988) Mutation Res., 204, 465-479). The potential confounders (smoking etc.), toxicokinetics of the agent(s) to be assessed and individual working practices should be carefully considered in any exposure assessment studies using human body fluid samples. Environmental monitoring on indicator cytostatics should be combined into studies designed to identify potential occupational exposure situations to anticancer agents. A properly performed study should also include dissemination of information to the workers to create a psychologically positive atmosphere for this important work.