Animal and epidemiological studies on the reproductive toxic effects of ethylene oxide (ETO) were considered in relation to occupational exposure levels (OELs) of ETO in the occupational environment of sterilisation units. Actual exposure levels in sterilisation units at Belgian and Dutch hospitals are presented and compared to data from recent studies conducted elsewhere. The animal studies did not match the actual exposure situation, involving a pattern of high peak levels and low time-weighted average levels. This may be the reason why epidemiological studies show contrasting results; they suggest reproductive toxicity of ETO at actual exposure levels. However, human data are scarce. There is a need for animal studies with a design that reflects the actual exposure situation. Epidemiological studies on reproductive events are also needed and a multi-country study would seem to be a possible approach, provided that the study design and data collection method are standardised.