Discordances between minimal residual disease estimates obtained by different methods are a problem in childhood acute lymphoblastic leukemia. We aimed to optimize methods allowing the biological exploration of such discrepancies, i.e. the combination of flow-sorting of small immunophenotypically defined cell populations with subsequent analyses of leukemia-associated cytogenetic and molecular marker. The approaches described here optimize the use of the same tube of unfixed, antibody-stained BM cells for flow-sorting of small cell populations and subsequent exploratory FISH and PCR-based analyses.