The research presented in this dissertation consists of three parts: 1) a description of the design and function of a novel data-mining program, LTR_STRUC; 2) a survey of LTR retrotransposons in the rice genome; 3) a survey of LTR retrotransposons in the mouse genome. The algorithm used by LTR_STRUC differs at a fundamental conceptual level from that employed in BLAST-type, query-based programs and thus provides an alternative, complementary method of identifying LTR retrotransposons in necleotide sequence data. We combined LTR_STRUC and conventional techniques to thoroughly search of the rice and mouse genomes for LTR retrotransposons. In rice, we found 59 families (37 copia-like, 20 gypsy-like, 2 non-autonomous). In mouse, we found 20 (all gypsy-like). In both species, we were able to more than double the number of recognized LTR retrotransposon families, a testament to the efficacy of LTR_STRUC-supported retrotransposon surveys.