A hallmark of cancer is tumor cell heterogeneity, which results from combinations of multiple genetic and epigenetic alterations within an individual tumor. In contrast, we have recently discovered that most human colorectal cancers (CRCs) are composed of mixtures of phenotypically distinct tumor cells organized into a stem cell hierarchy that displays a striking resemblance to the healthy colonic epithelium. We showed that long-term regeneration potential of tumor cells is largely influenced by the position that they occupy within the tumor's hierarchy. To analyze the organization of CRCs without the constraints imposed by tumor cell transplantation experiments, we have developed a method that allows for the first time tracking and manipulating the fate of specific cell populations in whole human tumors. This technology is based on editing the genomes of primary human CRCs cultured in the form of tumor organoids using Zinc-Finger Nucleases to knock-in either lineage tracing or cell ablation alleles in genes that define colorectal cancer stem cells (CRC-SCs) or differentiated-like tumor cells. Edited tumor organoids generate CRCs in mice that reproduce the tumor of origin while carrying the desired genetic modifications. This technological advance opens the gate to perform classical genetic and developmental analysis in human tumors. We will exploit this advantage to address fundamental questions about the cell heterogeneity and organization of human CRCs that cannot be tackled through currently existing experimental approaches such as: Are CRC-SCs the only tumor cell population with long term regenerating potential? Can we cure CRC with anti-CRC-SC specific therapies? Will tumor cell plasticity contribute to the regeneration of the CRC-SC pool after therapy? Do quiescent-SCs regenerate CRC tumors after standard chemotherapy? Can we identify these cells? How do common genetic alterations in CRC influence the CRC hierarchy? Do they affect the stem cell phenotype?