Jackson-Cook et al. (American Journal of Human Genetics 37:1049-1061, 1985) predicted a high risk of Down syndrome (DS) children for parents carrying a double NOR on acrocentric chromosomes. Hassold et al. (Human Genetics 76:381-384, 1987) could not confirm Jackson-Cook et al.'s findings, thus casting doubts on their conclusions. We studied the NORs of 1) 60 parents of 30 unselected DS subjects; 2) 30 unselected healthy subjects without trisomic offspring, who asked for chromosome analysis; and 3) 100 slides randomly chosen among 1,000 prepared by routine standard techniques and belonging to subjects who were chromosomally normal. By applying rigorously established techniques and scoring criteria we found 4 subjects (6.7%) with strictly defined double NORs (dNORs) in the DS parents sample, 2 subjects (6.7%) in the first control sample, and 3 (3%) in the second control sample. No significant difference among the observed frequencies of dNORs in the 3 samples could be demonstrated. Therefore, our data do not support Jackson-Cook et al.'s statements on the association of dNOR carrier status with DS offspring and on a highly increased risk of meiotic nondisjunction of chromosome 21 for a dNOR carrier parent. A tentative interpretation of the apparently contrasting cytogenetic findings would indicate that sampling and assignment biases are the main causes of this discrepancy.