The use of electronic documents has become a trend due to the rapid and convenient search functions it provides. In the tremendous amount of literature, contents might be intentionally or unintentionally changed, unrestrained or simply wrong. For example, alternative characters might be intentionally used to avoid certain characters. And parts might be unintentionally left out or wrong when documents are hand copied or printed. Similar situations might appear in electronic documents, such as input and typesetting errors. Therefore, users should be careful to compare the electronic document with paper document for verification and avoid errors. This study explores the strengths and weaknesses of electronic full-text databases, and discusses matters requiring attention in their use. Scripta Sinica is introduced and issues with its "Wen Xin Diao Lung" are described, such as issues with catalog data, text and punctuation marks, mistaking text as annotation, and wrongly written characters. Academia Sinica claims that Scripta Sinica is the "most exact Chinese full-text database," but the number of errors cannot be referred to as few. This study proposes recommendations for various issues of electronic full-text databases, including naming, full-text and annotation, version, version related data that should be provided, correctness of input, functional requirements of the search platform, character variations, e-books, database development team members, predictions, and feedback, hoping to create a positive effect and inspire the establishment electronic full-text databases. Errors are inevitable in electronic full-text databases, hence researchers must remain aware when using the databases, carefully consider its contents, be familiar with the strengths and weaknesses of each document, and compare the electronic version with paper version, so as to utilize its strengths and avoid its weaknesses. This is the only way to reduce the number of errors produced from using electronic full-text databases. Only then can the researcher be called a good user who can use the database to benefit his/her research.