Cytogenetic and DNA flow cytometric analyses of leukemic cells from 2,184 children with newly diagnosed acute lymphoblastic leukemia (ALL) identified 27 cases (1.2%) that had a hypodiploid line with fewer than 45 chromosomes per cell. Had cytogenetic techniques been used alone, seven cases would have been missed, compared with five if only flow cytometry had been used. For comparative purposes, the 27 cases were divided into three groups: near-haploid (n = 10), hypodiploid 30-40 (n = 9), and hypodiploid 41-44 (n = 8). Blast cells from patients with near-haploid ALL lacked structural chromosomal abnormalities; showed nonrandom retention of two copies of chromosomes 8, 10, 14, 18, 21, and the sex chromosomes; and had a second leukemic line with exactly twice the number of chromosomes or DNA content. Karyotypic analysis of the hypodiploid 30-40 and hypodiploid 41-44 groups disclosed structural abnormalities in the stemline or sideline of most of the well-banded cases; those in the latter group were similar to findings in cases with 45 chromosomes. As in the near-haploid group, chromosome 21 and the sex chromosomes were preferentially retained in the hypodiploid 30-40 and 41-44 cases. Except for a slight excess of female patients in the near-haploid group and an older age at diagnosis in the hypodiploid 30-40 cases, there were no initial clinical features that distinguished these patients from the general ALL population. Despite intensive treatment and short follow-up, 17 of the 27 patients have relapsed. This study suggests that the poor treatment responsiveness of hypodiploid ALL is not limited to the more than 80% of the patients who have 45 chromosomes per leukemic cell and demonstrates that cytogenetic and flow cytometric analyses are complementary in the evaluation of children with ALL.