Aim: This study aimed to examine the usefulness of seed density as a predictor of seed migration in patients with prostate cancer who received brachytherapy using Iodine-125 loose seed. Methods: From May 2006 to April 2016, 320 patients with localized prostate cancer underwent transperineal brachytherapy using iodine-125 loose seeds. Among them, 202 (63.1%) patients received brachytherapy monotherapy and 118 (36.9%) received combined brachytherapy and external beam radiotherapy. Seed density was calculated using the following formula: seed density = implanted seed number/prostate volume. All patients underwent radiography of the chest, abdomen and pelvis, and computed tomography at 1 day, 1 month, and 1 year after brachytherapy to evaluate the presence of seed migration. Results: In total, the number of implanted seeds was 21,876. Seed migration was detected in 92 (28.8%) patients. Of a total of 21,876 seeds, 144 (0.66%) showed migration. The number of needles, number of seeds, and seed density were significantly higher in the group with migration than in the group without migration (p = 0.05). The ROC cutoff values for prostate volume, number of needles, number of seeds, and seed density were 20.9 cc, 21, 65, and 3.0, respectively. In the univariate analysis, prostate volume, number of needles, number of seeds, seed density, and treatment modality were all significant factors in predicting migration (p = 0.05). In the multivariate analysis, seed density and treatment modality were significant factors in predicting migration (p = 0.05). Conclusion: Seed density is useful for predicting seed migration. In cases with seed density > 3.0, it is necessary to take measures such as considering the use of stranded seeds.