Proteomics has the potential to provide answers in cancer pathogenesis and to direct targeted therapy through the comprehensive analysis of protein expression levels and activation status. The realization of this potential requires the development of new, rapid, high-throughput technologies for performing protein arrays on patient samples, as well as novel analytic techniques to interpret them. Herein, we describe the validation and robustness of using reverse phase protein arrays (RPPA) for the analysis of primary acute myelogenous leukemia samples as well as leukemic and normal stem cells. In this report, we show that array printing, detection, amplification, and staining precision are very high, reproducible, and that they correlate with traditional Western blotting. Using replicates of the same sample on the same and/or separate arrays, or using separate protein samples prepared from the same starting sample, the intra- and interarray reproducibility was extremely high. No statistically significant difference in protein signal intensities could be detected within the array setups. The activation status (phosphorylation) was maintained in experiments testing delayed processing and preparation from multiple freeze-thawed samples. Differences in protein expression could reliably be detected in as few as three cell protein equivalents. RPPA prepared from rare populations of normal and leukemic stem cells were successfully done and showed differences from bulk populations of cells. Examples show how RPPAs are ideally suited for the large-scale analysis of target identification, validation, and drug discovery. In summary, RPPA is a highly reliable, reproducible, high-throughput system that allows for the rapid large-scale proteomic analysis of protein expression and phosphorylation state in primary acute myelogenous leukemia cells, cell lines, and in human stem cells. [Mol Cancer Ther 2006;5(10):2512–21]