This research discusses the analysis of sentiment in product reviews on E-Commerce Bukalapak using Natural Language Processing (NLP). The study aims to fill the knowledge gap regarding the analysis of product reviews in online stores in Indonesia, specifically Bukalapak. The data used in this research were collected from various product categories, such as clothing, electronics, cosmetics, and others. The method employed in this study was the TF-IDF method to train the Naive Bayes model. The results of the research show that the Naive Bayes model trained using the TF-IDF method achieved an accuracy of 88%. This indicates that the model has good capability in predicting the sentiment of product reviews. The analysis of positive reviews reveals customer satisfaction with product quality, fast delivery, reasonable pricing, and receiving items as expected. On the other hand, the analysis of negative reviews uncovers the mismatch between customer expectations and the actual conditions regarding color, delivery, and product orders. This study contributes to a deeper understanding of sentiment analysis in product reviews on E-Commerce Bukalapak. The insights from this analysis can be utilized by Bukalapak to enhance the quality of their products and services, providing a more satisfying experience for customers.