Selecting the right drug product is an important aspect in pharmaceutical services, especially for customers who do not yet have a deep understanding of the content and function of each drug. The limited number of pharmacists at Hero Farma Pharmacy often causes the service process to be inefficient, especially when still relying on manual recommendation methods that take a long time. This study aims to design a recommendation system that can assist in drug selection by implementing the content-based filtering method. This system is built by processing product attributes such as drug name, category, indication, dosage form, and price to form a profile of each product. The level of similarity between products is then calculated using the TF-IDF and Cosine Similarity methods. The data used in this study were obtained from Hero Farma Pharmacy located in Surakarta, with samples of 10 different types of drugs. The implementation results show that the system can produce recommendations with a good level of accuracy, where Promag drug obtained the highest Cosine Similarity value of 43.03% based on the query entered. This system successfully provides drug recommendations that have similar characteristics based on user needs and can help customers in making decisions correctly and improve the work efficiency of pharmacy staff.