Concerns with colloids, dispersed 1~1,000 nm particles, in the LILW repository are being raised due to their potential to enhance radionuclide release. Due to their large surface areas, radionuclides may sorb onto mobile colloids, and drift along with the colloidal transport, instead of being sorbed onto immobile surfaces. To prevent adverse implications on the safety of the repository, the colloidal impact must be evaluated. In this paper, colloid analysis done by SKB is studied, and factors to be considered for the safety assessment of colloids are analyzed. First, the colloid generation mechanism should be analyzed. In a cementitious repository, due to a highly alkaline environment, colloid formation from wastes may be promoted by the decomposition of organic materials, dissolution of inorganic materials, and corrosion of metals. Radiolysis is excluded when radionuclide inventory is moderate, as in the case of SKB. Second, colloid stability should be evaluated to determine whether colloids remain in dispersion. Stable colloids acquire electric charges, allowing particles to continuously repel one another to prevent coagulation. Thus, stability depends on the pH and ionic condition of the surroundings, and colloid composition. For instance, under a highly alkaline cementitious environment, colloids tend to be negatively charged, repelling each other, but Ca2+ ion from cement, acting as a coagulant, makes colloid unstable, promoting sedimentation. As in the case of SKB, the colloidal impact is assumed negligible in the silo, BMA, and BTF due to their extensive cement contents, but for BLA, with relatively less cement source, the colloidal impact is a potential concern. Third, colloid mobility should be assessed to appraise radionuclide release via colloid transport. The mobility depends on the density and size of colloids, and flow velocity to commence motion. As a part of the assessment, the filtration effect should also be included, which depends on pore size and structure. As in the case of SKB, due to static hydraulic conditions and engineering barriers, acting as efficient filters, colloidal transport is expected to be unlikely. In the domestic underground repository, the highly alkaline environment would lead to colloid formation, but due to high Ca2+ concentration and low flow velocity, colloids would achieve low stability and mobility, thus colloidal impact would be a minor concern. In the future, with further detailed analysis of each factor, waste composition, and disposal condition, reliable data for safety evaluation could be generated to be used as fundamental data for planning waste acceptance criteria.