The gold standard in the diagnosis of occupational asthma is the specific bronchial provocation test (sBPT), but other diagnostic criteria have been proven to have a similar sensitivity, mainly in asthma due to high molecular weight compounds. In order to assess wether some clinical findings can predict the positive response to sBPT, we studied 37 subjects (14 millers and 23 bakers) with suspected occupational asthma who underwent sBPT with wheat flour dust (dust exposure in a small cabin: geometric mean 12.1 mg/m3 for up to 30 min). A positive response to sBPT (FEV1 > 20%) was elicited in 20 subjects (11 early, 4 late, and 5 dual responses). There was no significant difference between subjects with positive or negative sBPT as regards mean age, smoking, length of employment, duration of symptoms, atopy (skin positivity to one or more common allergens) and PD20FEV1 methacholine. The percentage of subjects with work-related symptoms was significantly higher in subjects with positive sBPT with respect to subjects with negative sBPT (81% versus 41.2%, p < 0.01 by chi 2 test); furthermore, FEV1 was significantly lower in subjects with positive sBPT. The percentage of positive skin response to wheat flour extract (mean wheal diameter > or = 3 mm) was mildly but not significantly higher in subjects with positive sBPT (68.4% versus 41.2%). None of the following clinical factors (age < 35 years, asthma symptoms pre-existing occupational exposure, non smokers, atopy and bronchial hyperresponsiveness to methacholine), alone or in combination, were associated with higher prevalence of positive sBPT. We conclude that the response to sBPT in subjects with suspected occupational asthma due to flour dust can not be adequately predicted by other clinical, allergologic and functional data. Therefore, sBPT with flour dust should always be performed in subjects with suspected occupational asthma.