Minimal residual disease (MRD) predicts the outcome of acute lymphoblastic leukemia (ALL). Flow cytometry (FC) is one of the most sensitive and most applicable methods for MRD diagnostics, but there is still no agreement on the “gold standard” of the method. We tried to optimize flow cytometric MRD detection in T-ALL. Fourteen adults and 11 children with T-ALL and 12 normal bone marrow (BM) donors were enrolled in the study. We found that the most common phenotypic aberrations in T-ALL were TdT and CD99 coexpression on T-cells in BM. Therefore for MRD detection we developed a limited four-color marker panel (TdT/CD7/cCD3/CD19 and CD99/CD7/cCD3/CD2) and a standard analysis strategy. This assay was evaluated on BM of healthy controls. Less than 0.01% TdT+ or CD99 bright T-cells were found in normal BM. MRD was detected in 9 adult patients and 1 child at different time-points of treatment. The average TdT and CD99 mean fluorescence intensity (MFI) value of residual blasts fluctuated during therapy, but it still remained higher than MFI of normal T-cells. Our established MRD detection method differentiated leukemic lymphoblasts with sensitivity in the range of 0.01% and did not give any false positive results in normal BM.