Early clearance of leukemic cells is a favorable prognostic indicator in childhood acute lymphoblastic leukemia (ALL). However, identification of residual leukemic cells by their morphologic features is subjective and lacks sensitivity. To improve estimates of leukemia clearance, we applied flow cytometric techniques capable of detecting 1 leukemic cell in 10,000 or more normal cells and prospectively measured residual leukemia in bone marrow samples collected on day 19 of remission-induction chemotherapy from 248 children with newly diagnosed ALL. In 134 samples (54.0%), we identified at least 0.01% leukemic cells (0.01%-< 0.1% in 51 samples [20.6%], 0.1%-< 1% in 36 [14.5%], and > or = 1% in 47 [19.0%]). Among 110 children treated within a single chemotherapy program, the 5-year mean +/- SE cumulative incidence of relapse or failure to achieve remission was 32.2% +/- 6.5% for the 59 patients with 0.01% residual leukemic cells or greater on day 19 and 6.0% +/- 3.4% for the 51 patients with less than 0.01% leukemic cells (P <.001). The prognostic value of day-19 bone marrow status defined by flow cytometry was superior to that defined by morphologic studies and remained significant after adjustment for other clinical and biologic variables. Lack of detectable leukemic cells on day 19 was more closely associated with relapse-free survival than was lack of detectable residual disease at the end of remission induction (day 46). Thus, approximately half of the children with ALL achieve profound clearance of leukemic cells after 2 to 3 weeks of remission-induction chemotherapy, and these patients have an excellent treatment outcome.