morphism (12, 13). To gain insight into this interesting phenomenon, we analyzed the genomic organization of previously isolated tRNA genes (32). In the course of this study, we repeatedly identified the retrotransposable element DRE as well as the previously characterized transposable Tdd-3 element (30,38). Both elements were found associated with several unrelated tRNA genes in a characteristic position-specific manner.Up to now, DRE has never been found other than in association with tRNA genes in D. discoideum. In all analyzed clones, tRNA genes are separated by 50 ± 3 nucleotides from the associated DRE element, and DRE always occurs in a constant orientation relative to tRNA genes. Since no sequence similarity is apparent at the DRE integration site, a mechanism other than sequence-specific integration must be responsible for this striking position-specific and orientation-specific integration of DRE elements in D. discoideum. Similar results with respect to position specificity but not with respect to orientation specificity were obtained for Ty3, a retrotransposon identified in S. cerevisiae. Ty3 elements or a remnant LTR, termed sigma, were also found in front of various tRNA genes, separated in this case by 16 to 19 nucleotides from the mature coding tDNA sequence (10, 41). Both Ty3 and DRE, therefore, are members of a family of retrotransposons which integrate by a position-specific rather than by sequence-specific or random mechanisms into the chromosome (8, 31). Sequence-specific recombination has been described for many retrotransposable elements in different organisms (1,5,15,21,22,50), but many retrotransposons and most retroviruses are thought to integrate more or less at random (for reviews, see references 3, 42, and 45). In the cases of Ty3 and DRE, a great number and variety of tRNA genes mark the specific integration sites. This is a remarkable fact which puts these elements into a different category from other retrotransposons and