The aim of the study was to compare four different immunological methods to analyse fungal alpha-amylase in flour dust samples. Three European research groups have independently developed four immuno assay based methods to measure alpha-amylase in air samples. Three of the methods use polyclonal antibodies and one method uses monoclonal antibodies. Eighty personal samples have been collected during two to three work-shifts in four bakeries. Sampling was performed with PAS-6 inhalable dust samplers and aliquots from each sample were analysed by the three laboratories. The agreement between the four methods was high compared with agreement between immunological methods to measure other allergens in the air, e.g. for rat allergens. For the three methods with polyclonal antibodies the mean differences for individual samples was less than a factor of two. The arithmetic means (AM) of the estimated alpha-amylase exposure were 12.5, 11.3, 8.6 and 25.9ng/m(3) for the respective methods with values ranging from below the detection limit to 192, 215, 207 and 615 ng/m(3). The AM for all samples analysed by the methods with polyclonal antibodies varied by about a factor of 1.5. About one-third of the values were below or at the detection limit for all methods. In a regression analysis the squared correlation coefficients (R(2)) between the methods varied between 0.91 and 0.95 for the log transformed values. For workplace monitoring, results from the methods using polyclonal antibodies will be relatively comparable. High levels of alpha-amylase might differ in absolute numbers with a factor of two or more between the different methods but will anyway be considered as high and should result in preventive actions. On the other hand, this study also shows that despite the relative agreement between methods, there is a clear need for standardization.