The rapid growth of TikTok as a popular social media platform has significantly influenced written language practices, particularly in caption writing, which tends to be informal and spontaneous. This study aims to identify, classify, and analyze language errors found in TikTok captions created by Indonesian content creators. Employing a descriptive qualitative research design, the data were collected through documentation of publicly accessible TikTok captions and analyzed using qualitative content analysis. The findings reveal that language errors predominantly occur in the forms of redundancy, grammatical errors, spelling errors, and semantic inconsistency. These errors are not categorized as intentional code-switching or stylistic language mixing but rather as violations of standard linguistic norms. Theoretically, this study confirms that the informal nature of social media influences written language use and encourages non-standard forms in digital communication contexts.