Compositions that could be considered a precedent for aleatory composition date back to at least the late 15th century, with the genre of the catholicon, exemplified by the Missa cuiusvis toni of Johannes Ockeghem. A later genre was the Musikalisches Würfelspiel or musical dice game, popular in the late 18th and early 19th century. (One such dice game is attributed to Wolfgang Amadeus Mozart.) These games consisted of a sequence of musical measures, for which each measure had several possible versions and a procedure for selecting the precise sequence based on the throwing of a number of dice (Boehmer 1967, 9–47).
The French artist Marcel Duchamp composed two pieces between 1913 and 1915 based on chance operations. One of these, Erratum Musical written for three voices, was eventually published in 1934. Two of his contemporaries, Francis Picabia and Georges Ribemont-Dessaignes, also experimented with chance composition, these works being performed at a Festival Dada staged at the Salle Gaveau concert hall, Paris, on 26 May 1920. American composer John Cage's Music of Changes (1951) was "the first composition to be largely determined by random procedures" (Randel 2002, 17), though his indeterminacy is of a different order from Meyer-Eppler's concept. Cage later asked Duchamp: "How is it that you used chance operations when I was just being born?" (Lotringer 1998, ).
The earliest significant use of aleatory features is found in many of the compositions of American Charles Ives in the early 20th century. Henry Cowell adopted Ives's ideas during the 1930s, in such works as the Mosaic Quartet (String Quartet No. 3, 1934), which allows the players to arrange the fragments of music in a number of different possible sequences. Cowell also used specially devised notations to introduce variability into the performance of a work, sometimes instructing the performers to improvise a short passage or play ad libitum (Griffiths 2001). Later American composers, such as Alan Hovhaness (beginning with his Lousadzak of 1944) used procedures superficially similar to Cowell's, in which different short patterns with specified pitches and rhythm are assigned to several parts, with instructions that they be performed repeatedly at their own speed without coordination with the rest of the ensemble (Farach-Colton 2005). Some scholars regard the resultant blur as "hardly aleatory, since exact pitches are carefully controlled and any two performances will be substantially the same" (Rosner and Wolverton 2001) although, according to another writer, this technique is essentially the same as that later used by Witold Lutosławski (Fisher 2010) . Depending on the vehemence of the technique, Hovhaness's published scores annotate these sections variously, for example as “Free tempo / humming effect” (Hovhaness 1944, 3) and “Repeat and repeat ad lib, but not together” (Hovhaness 1958, 2).
In Europe, following the introduction of the expression "aleatory music" by Meyer-Eppler, the French composer Pierre Boulez was largely responsible for popularizing the term (Boulez 1957).
Other early European examples of aleatory music include Klavierstück XI (1956) by Karlheinz Stockhausen, which features 19 elements to be performed in a sequence to be determined in each case by the performer (Boehmer 1967, 72). A form of limited aleatory was used by Witold Lutosławski (beginning with Jeux Vénitiens in 1960–61) (Rae 2001), where extensive passages of pitches and rhythms are fully specified, but the rhythmic coordination of parts within the ensemble is subject to an element of chance.
There has been much confusion of the terms aleatory and indeterminate/chance music. One of Cage's pieces, HPSCHD, itself composed using chance procedures, uses music from Mozart's Musikalisches Würfelspiel, referred to above, as well as original music.