Purpose: To study the significance of Music Information Retrieval (MIR), the differences between text and music storage and retrieval, MIR systems’ tasks and other challenges.Methodology: Theoretical analysis.Results: There are different search engines and databases for MIR, like LastFM, Pandora and Muggle. Similarity, classification, transcription, rhythm and pitch recognition and Query-by-Humming (QBH) are the main tasks of MIR systems.Conclusion: Because of the differences between text and music qualities, text retrieval and MIR are different. To properly perform, each MIR system needs to consider: incorrect melody singing in QBH, difficulties of transcribing polyphonic music, genre’s dependency to several classes, the lack of clear sense among tags and similarities in various pieces of music.