The majority of tumors, including melanoma, are phenotypically heterogeneous in that they contain various cell populations with differential expression of cell surface antigens such as CD133/Prominin-1. We have used fluorescence-activated cell sorting (FACS) technology to purify CD133(+) and CD133(-) cellular subsets from mouse melanoma models for high-quality total RNA practical for downstream applications such as expression profiling. Implementation of this strategy can lead to higher resolution of transcripts that are potentially important for the survival and functionality of one cancer cell population relative to another. Suboptimal extraction of RNA after FACS is common and can ultimately result in misinterpretations that impede the effective design of novel therapies. Here, we describe a number of methods that have been amenable to the successful isolation of high-quality total RNA after FACS of CD133(+) and CD133(-) mouse melanoma cell fractions.