We used a direct polymerase chain reaction (PCR) method for quantification of HPRT exons 2 + 3 deletions and t(14;18) translocations as a measure of illegitimate V(D)J recombination. We determined the baseline frequencies of these two mutations in mononuclear leukocyte DNA from the umbilical cord blood of newborns and from the peripheral blood of adults. In an initial group of 21 newborns, no t(14;18) translocations were detected (< 0.049 x 10(-7)). The frequency of HPRT exons 2 + 3 deletions was 0.10 x 10(-7) per mononuclear leukocyte, lower than expected based on the T-cell proportion of this cell fraction (55%-70%) and previous results using the T-cell cloning assay (approximately 2-3 x 10(-7) per clonable T-cell). Phytohemagglutinin (PHA), as used in the T-cell cloning assay, was examined for its effect on the frequencies of these mutation events in mononuclear leukocytes from an additional 11 newborns and from 12 adults. There was no significant effect of PHA on t(14;18) translocations which were rare among the newborns (1 detected among 2.7 x 10(8) leukocytes analyzed), and which occurred at frequencies from < 1 x 10(-7) (undetected) to 1.6 x 10(-4) among the adults. The extremely high frequencies of t(14;18)-bearing cells in three adults were due mainly to in vivo expansion of two to six clones. However, PHA appeared to stimulate a modest (although not significant) increase in the frequency of HPRT exons 2 + 3 deletions in the leukocytes of the newborns, from 0.07 x 10(-7) to 0.23 x 10(-7). We show that both the direct PCR assay and the T-cell cloning assay detect similar frequencies of HPRT exons 2 + 3 deletions when calculations are normalized to blood volume, indicating that the apparent discrepancy is probably due to the different population of cells used in the assays. This direct PCR assay may have utility in characterizing the effects of environmental genotoxic agents on this clinically important recombination mechanism.