The present review is based on findings from 178 publications retrieved through an extensive search of the MedLine/PubMed database for a 25 years time period (1980-2004) and 10 manually identified papers. Among the cytogenetic biomarkers that are frequently used in field studies, chromosome aberrations (CA) and micronuclei (MN) but not sister chromatid exchanges (SCE) were found consistently increased in children exposed to environmental pollutants. Meta-analysis of the studies reporting SCE in cord blood showed similar levels of SCE in exposed and in non-exposed newborns. Exposure to airborne pollutants, soil and drinking water contaminants, mostly increased CA and, to a lesser extent, MN levels in children. The effect of exposure to airborne urban pollutants was consistently reported by field studies measuring DNA, albumin and hemoglobin adducts. Prenatal (in utero) and postnatal exposure (environmental tobacco smoke, ETS) to tobacco smoke compounds were associated with increased frequencies of DNA and hemoglobin adducts and CA. The limited number of field studies measuring DNA fragmentation (Comet assay), hypoxanthine-guanine phosphoribosyltransferase (HPRT) and the glycophorinA (GPA) mutation frequency in environmentally exposed children precluded a meaningful evaluation of the usefulness of these assays. Meta-analyses performed in children exposed to ETS and in newborns exposed in utero to their mothers' smoke showed 1.3 and 7 times higher levels of hemoglobin adducts compared to referent subjects, respectively. These increases are consistent with the epidemiological evidence of higher lung cancer risks reported in adults who had never smoked and were exposed to ETS during childhood and with 7-15 times higher lung cancer risks reported in smokers than in non-smokers. Higher levels of PAH-DNA adducts were found in fetal than in maternal tissue, suggesting a specific susceptibility of the fetus to this class of ubiquitous environmental pollutants. According to these findings, future research and biomonitoring programs on children would greatly benefit from the inclusion of selected biomarkers that could provide biologically based evidence for the identification of intervention priorities in environmental health.