Before market introduction, genetic engineered (GE) food products, like any other novel food product, are subjected to extensive assessment of their potential effects on human health. In recent years, a number of profiling technologies have been explored aiming to increase the probability of detecting any unpredictable unintended effect and, consequently improving the efficiency of GE food safety assessment. These techniques still present limitations associated with the interpretation of the observed differences with respect to their biological relevance and toxicological significance. In order to address this issue, in this study, we have performed 2D-gel electrophoresis of five different ears of five different MON810 maize plants and of other five of the non-transgenic near-isogenic line. We have also performed 2D-gel electrophoresis of the pool of the five protein extractions of MON810 and control lines. We have notice that, in this example, the exclusive use of data from 2D-electrophoresed pooled samples, to compare these two lines, would be insufficient for an adequate safety evaluation. We conclude that, when using "omics" technologies, it is extremely important to eliminate all potential differences due to factors not related to the ones under study, and to understand the role of natural plant-to-plant variability in the encountered differences. (C) 2010 Elsevier Inc. All rights reserved.