Rice is a model genome for cereal research, providing important information about genome structure and evolution. Retrotransposons are common components of grass genomes, showing activity at transcription, translation and integration levels. Their abundance and ability to transpose make them good potential markers. In this study, we used 2 multilocus PCR-based techniques that detect retrotransposon integration events in the genome: IRAP (inter-retrotransposon amplified polymorphism) and REMAP (retrotransposon-microsatellite amplified polymorphism). Markers derived from Tos17, a copia-like endogenous retrotransposon of rice, were used to identify genetic similarity among 51 rice cultivars (Oryza sativa L.). Genetic similarity analysis was performed by means of the Dice coefficient, and dendrograms were developed by using the average linkage distance method. A cophenetic correlation coefficient was also calculated. The clustering techniques revealed a good adjustment between matrices, with correlation coefficients of 0.74 and 0.80, or lower (0.21) but still significant, between IRAP and REMAP-based techniques. Consistent clusters were found for Japanese genotypes, while a subgroup clustered the irrigated Brazilian genotypes.