Parental age is the most important aetiological factor in trisomy formation in humans. Cytogenetic studies on germ cells reviewed here imply that (i) 2-4% sperm are aneuploid and 8.6% oocytes from IVF are hyperploid (ii) a paternal age effect may exist, and (iii) oocytes of aged women contain precociously separated chromatids in metaphase II. Trisomy data suggest that most aneuploidy is generated during meiosis I of oogenesis and is maternal age-dependent. Trisomy 18 is unique, originating mostly from maternal meiosis II errors. The extra gonosome in 47, XXY derives mostly from a paternal meiosis I error. Trisomy of individual chromosomes may remain low, linearly rise, or exponentially increase with advanced maternal age. Maternal age related trisomies involve achiasmatic and normochiasmate chromosomes, and chromosomes with disturbed recombination and distally located chiasmata. Hypotheses on the origin of the maternal age effect are critically reviewed. One model is presented that relates to altered cell cycle and protein phosphorylation in oocytes of aged mammals and accounts for most of the observed data in humans and in experimental studies. Aneuploidy may thus involve a predetermined component but is possibly also influenced by extrinsic factors reducing oocyte quality or depleting the oocyte pool precociously. Areas of future research are proposed to elucidate (i) the significance of early disturbances in the prenatal ovary, (ii) parameters diminishing the quality of oocytes in dictyate stage, and (iii) mechanisms enabling oocytes to process all chromosomal configurations successfully during later stages of oogenesis. Studies with newly developed and existing animal models appear indispensable to identify exposures affecting chromosome disjunction during meiosis, especially in the aging female.