Initial research revealed interaction effects on health measures between exposure to neurotoxicants and age. Symptom reports of workers were conspicuously increased if high-concentration occupational exposure (e.g. to organic solvents, lead) was combined with age above 54 years. The symptom increase in elderly workers was interpreted as a possible indicator of a biological vulnerable phase or delayed response of former high exposure. A second study of the hypothesized age-exposure interaction was performed with a group of workers who had homogenous exposure to a single organic solvent using a neurobehavioral performance evaluation (the EURO-NES). Workers in the rotogravure printing industry who were exposed to toluene were examined two times with an interval of one year (n =333/278). The sample was stratified by workers with significantly different toluene exposure, printers and end-processing operators, and four age classes (< 31, 31-40, 41-50, > 50). The mean lifetime weighted average exposure (LWAE) varied depending on age classes and years of employment with exposure between 7 and 17 ppm in the operators and between 35 to 62 ppm toluene in the printers. Multivariate analyses revealed a significant performance decrease with age (simple reaction time, symbol digit, switching attention, digit span). Again an interaction between age and exposure was found depending on diverging psychometric performance trends with older age. However, contrary to the hypothesis the group with higher exposures (printers) and older age revealed better performance and less symptoms than the group with lower exposure (end-processing operators). The paradoxical results are explained by differences in the intellectual capability in the oldest strata and a possible reversibility of neurobehavioral effects of former high toluene exposure under the condition of later low exposure. There are no hints of adverse delayed effects of former toluene exposure in a possible vulnerable phase in age over 50 years. The different interaction findings of the initial and present study seem to depend mainly on exposure differences in quality and quantity.