Two of the four repeated DNA sequences near the 5' end of the silk fibroin gene hybridize with discrete-length families of repeated DNA. These two families comprise 0.5% of the animal's genome. A repeated sequence with a conserved length has also been-found in the short class of moderately repeated sequences in the sea urchin. The discrete length, interspersion, and sequence fidelity ofthese moderately repeated sequences suggests that each has been multiplied as a discrete unit. Thus, transposition mechanisms may be responsible for the multiplication and dispersion of a large class of repeated sequences in phylogenetically diverse eukaryotic genomes. The repeat we have studied in most detail differs from previously described eukaryotic transposable elements: it is much shorter (1300 base pairs) and does not have terminal repetitions detectable by DNA hybridization. A simple technique for identifying such discrete-length repeated sequences is described, Repeated DNA sequences comprise 20-50% of most animal genomes (1) and >50% of many plant genomes (2). Detailed information has become available on the organization, distribution, and evolution of many satellite and repeated structural gene DNAs, but few details are known about the arrangement of the bulk of the moderately repeated, interspersed DNA sequences present in most eukaryotes. Recent work that links genetic and structural studies on.repeated sequences in yeast (3,4) and Drosophila (5-7) indicates that some eukaryotic repeated sequences are transposable elements analogous to those found in prokaryotes. These eukaryotic transposable elements are >5000 nucleotides (Nt) long, much longer than the short repeated sequences found throughout the metazoa (1).In most eukaryotes, and in silk moths and sea urchins in particular, repeated sequences that average -400 Nt are interspersed with single-copy DNA averaging 1000-3000 Nt. In contrast, some animals, including Drosophila, seem to have few short repeated sequences interspersed with nonrepeated DNA (8-10). Our studies have been concerned with repeated sequence organization in genomes that have a repeat interspersion typical of most animals (1, 11-13).The length ofa repeated sequencerefers to the duplex region formed when a repeated sequence renatures. The unreacted single-strand tails mark the ends. These single-strand tails have been visualized with the electron microscope (14) or degraded with nuclease S1 and the remaining duplexes were fractionated on the basis of length (13). We have used Southern transfer hybridization of populations of nuclease Sl-trimmed repeated sequences to examine the distribution of lengths of individual members of various repeated-sequence families. We find that some repeated-sequence families consist of members that have the same length, while other families have members that differ in length (15). Three discrete-length repeated sequences are described, two in the DNA of the silk moth Bombyx moti near the 5' end of the silk fibroin gene (15)