Down syndrome caused by meiotic nondisjunction of chromosome 21 in humans, is well known to be associated with advanced maternal age, but success in identifying and understanding other risk factors has been limited. Recently published work in a U.S. population suggested intriguing interactions between the maternal age effect and altered recombination patterns during meiosis, but some of the results were counter-intuitive. We have tested these hypotheses in a population sample from India, and found that essentially all of the results of the U.S. study are replicated even in our ethnically very different population. We examined meiotic recombination patterns in a total of 138 families from the eastern part of India, each with a single free trisomy 21 child. We genotyped each family with a set of STR markers using PCR and characterized the stage of origin of nondisjunction and the recombination pattern of maternal chromosome 21 during oogenesis. Our sample contains 107 maternal meiosis I errors and 31 maternal meiosis II errors and we subsequently stratified them with respect to maternal age and the number of detectable crossover events. We observed an association between meiosis I nondisjuncion and recombination in the telomeric 5.1 Mb of chromosome 21. By contrast, in meiosis II cases we observed preferential pericentromeric exchanges covering the proximal 5.7 Mb region, with interaction between maternal age and the location of the crossover. Overall reduction of recombination irrespective of maternal age is also evident in meiosis I cases. Our findings are very consistent with previously reported data in a U.S. population and our results are the first independent confirmation of those previous reports. This not only provides much needed confirmation of previous results, but it suggests that the genetic etiology underlying the occurrence of trisomy 21 may be similar across human populations. © 2009 Wiley-Liss, Inc.