A challenge modern molecular biology is faced with, is defining the organisation of intracellular signalling networks. The latest advances in microarrays and high throughput proteomics allow us to study changes in transcription, translation and post-translational modification at the level of the whole genome, potentially leading to the identification of all genes affected by a given stimulus. These approaches are very robust, however, no direct information can be obtained about the signalling events leading to those observed changes. We liken this situation to finding and studying eggs laid by a group of hens, but having no idea which chicken laid the eggs of interest. A complementary strategy is to screen for signalling components (“chickens”), whose activation leads to changes in gene expression (“eggs”). The most common example of this approach is a genetic screen, where mutations causing particular phenotypes are identified. These screens are mainly applied in organisms wi th a relatively short generation time such as bacteria, yeast, the worm and the fly. In these systems however, detectable phenotypes only occur for non-redundant signalling molecules. Thus even using all existing screening techniques including knockout, genome-sequencing projects reveal a large number of novel genes with no previously identified function, even for the best-studied organisms. In addition, the possible involvement of the majority of those gene products with assigned function in complex processes, such as embryogenesis or inflammation is unknown. It is therefore desirable to design a high throughput experimental system, which detects gene products based on their function. One version of such a system has been developed recently and is being successfully used for identifying components of the IL-1 / TNF signalling network, utilizing a transcription reporter sensitive to these cytokines. Upon overexpression, most of the known components of this signalling network mimic the effect of the cytokine. Based on this observation, bioactive protein-encoding cDNA clones can be identified in complex pools by the sensitive reporter. This system provides a proof of principle for a general strategy, whereby gene products are being detected based on their function.