Sufficient information is now available from the literature to produce an audit of trisomy 16, in a theoretical cohort of 100,000 recognized pregnancies, from gametogenesis to term and onwards. Recent reports of premature separation of chromosome 16 bivalents during maternal meiosis I provide a novel mechanism for generation of this aneuploidy. Most, if not all, errors resulting in recognized mosaic and non-mosaic trisomy 16 pregnancies investigated using polymorphic DNA markers appear to originate at that stage. The incidence of this maternally derived trisomy 16 in the late first trimester is equivalent to 1500 cases in 100,000 recognized pregnancies, a figure which now corresponds very closely to the reinterpreted oogenesis data. Most trisomy 16 pregnancies are lost around 12 weeks' gestation, but of the order of 10 per cent (120-150 in this audit) undergo reduction to disomy, with 30 of these excluding aneuploidy from the fetal cell lineage (trisomic zygote rescue) and continuing into the second trimester. Maternal uniparental disomy (UPD) in one-third of this latter group is associated with loss later in pregnancy or severe intrauterine growth retardation, but can be compatible with a viable pregnancy. Adverse pregnancy outcomes are not restricted to those with UPD. Analysis of reports of confined placental mosaicism for chromosome 16 without associated UPD indicates that the presence of high levels of trisomic cells in the placenta alone consistently produces a more variable inhibition of fetal growth, which may also, in cases, be associated with late pregnancy loss.