Cancer is a genetic disease that ultimately results from the failure of cells to respond correctly to diverse signals. Signal transduction and signal integration are highly complex, requiring the combinatorial interaction of multiple genes. Classical genetics in model organisms including Caenorhabditis elegans has been of immense use in identifying nonredundant components of conserved signalling pathways. However, it is likely that there is much functional redundancy in the informational processing machinery of metazoan cells; we therefore need to develop methods for uncovering such redundant functions in model organisms if we are to use them to understand complex gene interactions and oncogene cooperation. RNAi may provide a powerful tool to probe redundancy in informational networks. In this review, I set out some of the progress made so far by classical genetics in understanding redundancy in gene networks, and outline how RNAi may allow us to approach this problem more systematically in C. elegans. In particular, I discuss the use of genome-wide RNAi screens in C. elegans to identify synthetic lethal interactions and compare this with synthetic lethal interaction analysis in Saccharomyces cerevisiae.