Beer was chosen as a cereal-derived and homogeneous product for a comparison of organic and conventional production methods in terms of mycotoxin contamination levels. Ochratoxin A (OTA, a storage mycotoxin) and deoxynivalenol (DON, a field mycotoxin) were assessed by HPLC in organically and conventionally produced beers sold in Belgium. Immunoaffinity column (OchraTest and DONPrep) purification was used prior to HPLC analysis. For in-house validation, recovery experiments, carried out with the spiked beers in the ranges of 50-200 ng OTA l-1 and 20-100 microg DON l-1, led to the overall averages of 91% (RSD = 10%, n = 9) and 93% (RSD = 5%, n = 27), respectively. Organic beers collected during 2003-2004 were more frequently OTA-contaminated (95%, n = 40) than their conventional counterparts (50%, n = 40). Conventional beers were OTA-contaminated at a mean concentration of 25 ng l-1 (range: 19-198 ng l-1), while organic beers contained a mean level of 182 ng l-1 (range: 18-1134 ng l-1). High OTA contamination above the limit of 200 ng l-1 (up to 1134 ng l-1) occasionally occurred in organically produced beers. A complementary survey performed with the same brands in 2005 did not confirm this accidental presence of excessive OTA loads (range: 3-67 ng l-1 for 10 conventional beers and 19-158 ng l-1 for 10 organic beers). Establishing a maximum of 3 microg OTA kg-1 in malt, the application of the regulation EC No. 466/2001 (entered in force before the last sampling) may be related to the observed improvement. The overall incidence of DON was 67 and 80% in conventional and organic beers, respectively. DON concentrations ranged from 2 to 22 microg DON l-1 (mean = 6 microg DON l-1) in conventional beers, while organic beers ranged from 2 to 14 microg DON l-1 (mean=4 microg DON l-1). Thus, DON in beers does not appear to be a major matter of concern. From the statistical tests, it was concluded that the variation between different batches was significant (P < 0.0001), in contrast to that observed between different brands, showing a lack of homogeneity in the raw materials. This occurs either in organically or in conventionally produced materials. Considering these results, an optimized frequency of controls according to European Regulations EC No 466/2001 and EC No 856/2005 should be recommended to reject the irregular batches.