Concerns have been raised that genomewide selection may hasten the loss of genetic diversity in plant breeding programs. Our objective was to determine if genomewide selection and phenotypic selection lead to different levels of genetic similarity among the selected lines within a biparental population. The best 5, 10, 25, and 50% of F3 lines within each of 27 maize (Zea mays L.) biparental populations were identified by phenotypic selection for testcross performance and by genomewide selection. Without any selection, the mean genetic similarity among lines within a population was 0.52 and the maximum similarity among lines was 0.71. Averaged across the 27 populations, phenotypic selection for grain yield, moisture, and an index of these two traits did not cause a significant increase in genetic similarity among the selected lines. Likewise, genomewide selection of the best 50% of the lines did not lead to a significant increase in genetic similarity. In contrast, genomewide selection of the best 5% of lines significantly increased the mean similarity from 0.52 to 0.57. For comparison, the mean similarity was 0.68 among the 5% most similar lines. The minimal increase in genetic similarity with genomewide selection was attributed to the absence of lines with the perfect or near-perfect marker profile in the biparental populations. The general level of genetic diversity within the training population had no effect on the genetic similarity among the selected lines. We conclude that genomewide selection causes only a minimal loss of genetic diversity within biparental populations.