One of the advantages of having information in digital form is that it lends itself readily to content-based access. This applies to information stored in any media, though content searching through information stored in a structured database or as text is more developed than content searching through information stored in other media such as music In practice, the most common way to index and provide retrieval on digital music is to use its metadata such as title, performer, etc , as has been done in Napster. My research has lead to the development of a digital music information retrieval system called Ceolaire which can index monophonic music files Music files are analysed for notes on the equal tempering scale, where note changes are observed and recorded as being up (U), down (D) or the same (S) relative to the previous note. These note changes are then indexed in a search engine At query time, notes are generated by a user using a web based interface These notes form the query for the retneval engine. A user is presented with a ranked list of highly scored documents. This thesis explores the building and evaluation of the Ceolaire system.