Abstract A qualitative screening method was validated for the detection of genetically modified organisms (GMOs) in various processed food matrices (cooked maize grit, infant formula, biscuits, meal of acidified soybeans). The prepared food matrices contained each 0%, 2%, 100% (10% instead of 100% in the case of biscuits) of Roundup-Ready© soybeans and/or of Bt-176 maize. The method was based on the detection of the introduced DNA sequences by using the polymerase chain reaction (PCR) for amplification of DNA. Two different detection systems were applied: one based on the 35S promoter fragment which is present in GM-soybeans and in GM-maize and one based on the nos terminator sequence which is present only in GM-soybeans. Prior to the validation study, it was demonstrated that the food matrices were homogenous and that there was no cross-contamination between GMO-containing and GMO-free samples. Some laboratories had a high portion of false positive results probably due to laboratory contamination whereas all others reported nearly always correct data for all matrices investigated. Statistical analysis showed that after exclusion of the outliers, an average of 97% correct results was obtained for non-GMO containing samples (3% false positive results) and an average of 98% correct results for GMO-containing samples (2% false negative results). The validated method was shown to be suitable for screening of GMOs in processed food matrices also when extreme physical stress was applied (e.g., heating for 45 min at 100 °C or 10 min at 180 °C) or when multiple components were present, as in the case of biscuit preparations.