Background: Rare cancer stem cells (CSC), proposed to be solely responsible for tumor propagation and re-initiation, are functionally identified as tumor-initiating cells (TIC) from ex vivo tumors using xenotransplantation and clonogenic limiting dilution assays (LDA). TIC have not previously been described from ex vivohuman clear cell renal cell carcinoma (ccRCC). Methods: Primary human ccRCC samples (n=120) from patients undergoing nephrectomy were processed and implanted as subcapsular fragments or cell suspension injection LDAs with Matrigel in NOD/SCID/IL2Rγ^-/- (NSG) mice, and observed for at least 6 months. In vitro clonogenic LDAs assays were performed from primary cell suspensions and ccRCC cell lines. LDAs were supplemented with human stromal cells and proteins, and the Y-26732 ROCK inhibitor. Multiparametric flow cytometry and immunofluorescence were used to investigate tumor heterogeneity and cell viability. Results: ccRCC TIC appeared rare from injected suspensions, but xenografts engrafted frequently from tiny fragments, and clonogenic frequencies were 10³-10⁴greater than TIC frequencies, suggesting that LDAs underestimated ccRCC tumor cell potential. We systematically identified multiple methodological steps that distort quantitation and identification of ccRCC TIC. For example cell viability was highly variable prior to processing, disaggregation itself destroyed up to 99% of tumor cells, standard assays substantially overestimated tumor cell viability in suspensions, and supplementation with human extracellular cells or proteins, or inhibition of anoikis by Y-26732 increased clonogenic and TIC frequencies in cell lines and primary ccRCC suspensions. Annexin-V staining revealed that tumor cells were more apoptotic then normal stromal cells, and that tumor cells positive for CD44 (a putative CSC marker) were more viable than CD44- tumor cells. Conclusions: We describe multiple, unappreciated and largely unavoidable observational errors in essential methods used to study TIC in ccRCC. ccRCC TIC may be more common than appreciated. Re-examination of the CSC hypothesis in other solid tumors is warranted in view of these previously unexplored methodological biases.