Twenty-four cases of trisomy 13 and one case with disomy 13, but a de novo dic(13,13) (p12p12) chromosome, were examined with molecular markers to determine the origin of the extra (or rearranged) chromosome. Twenty-one of 23 informative patients were consistent with a maternal origin of the extra chromosome. Lack of a third allele at any locus in both paternal origin cases indicate a somatic duplication of the paternal chromosome occurred. Five cases had translocation trisomy: one de novo rob(13q14q), one paternally derived rob(13q14q), two de novo t(13q13q), and one mosaic de novo t(13q13q)/r(13). The patient with a paternal rob(13q14q) had a maternal meiotic origin of the trisomy; thus, the paternal inheritance of the translocation chromosome was purely coincidental. Since there is not a significantly increased risk for unbalanced offspring of a t(13q14q) carrier and most trisomies are maternal in origin, this result should not be surprising; however, it illustrates that one cannot infer the origin of translocation trisomy based on parental origin of the translocation. Lack of a third allele at any locus in one of the three t(13q13q) cases indicates that it was most likely an isochromosome of postmeiotic origin, whereas the other two cases showed evidence of recombination. One balanced (nontrisomic) case with a nonmosaic 45, -13, -13, +t(13;13) karyotype was also investigated and was determined to be a somatic Robertsonian translocation between the maternal and paternal homologues, as has been found for all balanced homologous Robertsonian translocations so far investigated. Thus, it is also incorrect to assume in de novo translocation cases that the two involved chromosomes are even from the same parent. Despite a maternal origin of the trisomy, we cannot therefore infer anything about the parental origin of the chromosomes 13 and 14 involved in the translocation in the de novo t(13q14q) case nor for the two t(13;13) chromosomes showing a meiotic origin of the trisomy.