microRNAs (miRNAs) are short, noncoding RNAs that negatively regulate the levels of mRNA post-transcriptionally. Recent experiments revealed thousands of mRNA-miRNA pairs in which multiple miRNAs may bind the same transcript. These results raised the notion of miRNAs teamwork for a wide range of cellular context. miRror2.0 utilizes the miRNA-target predictions from over a dozen programs and resources and unifies them under a common statistical basis. The platform, called miRror2.0, considers the combinatorial regulation by miRNAs in different tissues, cell lines and under a broad range of conditions. A flexible setting permits the selection of the preferred combination of miRNA-target prediction resources as well as the statistical parameters for the analysis. miRror2.0 covers six major model organisms including human and mouse. Importantly, the system is capable of analyzing hundreds of genes that were subjected to miRNAs' regulation. Activating miRror2.0 by introducing thousands of genes from miRNA overexpression experiments successfully identified the objective miRNAs. The output from miRror2.0 is a list of genes that is optimally regulated by a defined set of miRNAs. A symmetric application of miRror2.0 starts with a set of miRNAs, and the system then seeks the preferred set of genes that are regulated by that miRNA composition. The results from miRror2.0 are empowered by an iterative procedure called PSI-miRror. PSI-miRror tests the robustness of miRror2.0 prediction. It allows a refinement of the initial list of genes in view of the miRNAs that optimally regulate this list. We present miRror2.0 as a valuable resource for supporting cellular experimentalists that seek recovery of combinatorial regulation by miRNAs from noisy experimental data. miRror2.0 is available at http://www.mirrorsuite.cs.huji.ac.il .