The development of information technology has encouraged the digitalization of various forms of information, including regional language dictionaries. The Mandailing Language Dictionary in digital form makes it easier for people to learn and preserve regional languages. However, digital data storage also poses risks to data security, such as theft, duplication, or modification of data by unauthorized parties. Therefore, a method is needed to improve data security in digital dictionary systems. This study aims to implement the Least Significant Bit (LSB) method as a steganographic technique to hide text data within digital images in the Mandailing Language Dictionary system. The LSB method works by embedding text data into the least significant bit of digital image pixels so that the existence of the data is not easily detected visually. This research produces a web-based system capable of performing encoding (data embedding) and decoding (data extraction) processes on digital images. The encoding process is carried out by converting text into binary form and embedding it into the red color channel of each image pixel using the LSB method. In the decoding process, the system reads the LSB bits from the image to retrieve the embedded text data. The testing results show that the embedded data can be accurately extracted without any changes, and the image quality before and after the embedding process does not show significant differences. Therefore, the LSB method can be used as a solution to improve data security in the Mandailing Language Dictionary system.