Structural, and numeric chromosome aberrations (CA), sister-chromatid exchange (SCE), phytohemagglutinin stimulation (LI), proliferative rate index (PRI), and UV light-induced unscheduled DNA-synthesis (UDS) were investigated in peripheral blood lymphocytes (PBL) of 48 historical controls ("Controls"); of 14 hospital controls in Budapest, Hungary ("Budapest controls"); of 9 nurses occupationally exposed to low-dose ethylene-oxide in Budapest ("Budapest exposed"); of 10 hospital controls in Eger, Hungary ("Eger controls"); and of 27 high dose ETO-exposed nurses in Eger ("Eger exposed"), where neoplasias, mainly breast cancers, were observed. ETO concentrations in the ambient air samples varied from 5 to 20 mg/m3 in Budapest; and from 5 to 100 mg/m3 in Eger. Bothe LI and PRI were depressed in Budapest exposed, indicating ETO-induced cytotoxicity and, however, normal in Eger exposed. SCE was slightly elevated in Budapest exposed, but significantly increased in Eger exposed. The yields of cells with high frequency SCE (HFC) were only increased in Eger exposed. The expected low CA frequencies were found in Controls and in Budapest controls. ETO exposures significantly increased the CA frequencies in Budapest and Eger exposed. In Budapest exposed, as expected, we found deletions; in a lesser extent chromatid exchanges and dicentrics; but no rings were detected. These results are in a good accordance to the published data of other investigations carried out on ETO-exposed human populations. However, in Eger exposed, beside the increased yields of deletions, the frequencies of dicentrics and rings showed a significant excess compared to the reviewed data. An unexpected, significant increase of dicentric and ring frequencies was also detected among the hospital controls in Eger controls without known clastogenic exposure. The role of confounding factors (age, smoking and drinking habit, total leukocyte count and hematocrit) was investigated by an analysis of variance on CA and SCE frequencies in Controls and in Eger exposed. Leukocyte count and mean age showed only significant effects on CA in Eger exposed and on SCE in Controls, respectively. A possible active confounding factor could be the temporary natural radioactivity of the local tap water. UDS in Budapest exposed and in Eger control were significantly higher then in the Controls and in Budapest controls. In Eger exposed UDS was significantly decreased compared to the Budapest exposed and Eger control groups. The explanation of the present results is difficult on the basis of the reviewed data on ETO-induced CA frequencies in exposed human populations, and it raises an issue of an independent genotoxic effect in Eger, which is common both in Eger controls and in Eger exposed, such as natural radioactivity.