The music industry is currently growing very rapidly. The limited time possessed by consumers to consume products conventionally is the main reason for online purchases by consumers. This change also occurred in the music industry, so that online music sites that provide convenience in service and this convenience made consumers decide to buy music products online and abandoned the old habit of consuming conventional music products on CDs. The ease of consuming songs online does not make the user easy to determine the songs they want to hear. The difficulty of users in determining what songs to listen to in a web music can be overcome by the existence of a music recommendation system, where music sites will provide recommendations on songs based on user preferences. This study aims to create a music recommendation system on online music sites to provide convenience for consumers in choosing music products according to their preferences. This study explains the problem of users of online music sites who do not rank a song that has been heard, and overcome it with a recommendation system using user-based collaborative filtering methods.