In the Roman liturgy adopted in Francia in the mid-eighth century, the tract Qui habitat was sung within the Good Friday liturgy. This was retained in Rome until the suppression of Old Roman chant in the thirteenth century. In the Romano-Frankish tradition, however, it was replaced at some point by a newly composed second-mode tract, Eripe me. In the first half of the tenth century, this chant was identified as ‘nuperrime compilatum’ in De Divinis Officiis (Pseudo-Alcuin). However, the composition and incorporation into the liturgy of Eripe me considerably predates this identification.

The Roman practice of singing Qui habitat on Good Friday is reflected in several of the Ordines Romani. Since these Frankish redactions of Roman liturgical practice often retain archaisms, the presence of Qui habitat in a manuscript does not necessarily mean either that it predates the composition of Eripe me, or that it predates the adoption of Eripe me in a particular locale. Graduals and Cantatoria are more likely to reflect current practice rather than to preserve archaisms, since they had a practical function. To retain a rubric referring to Qui habitat while copying an Ordo Romanus would not affect daily practice at the institution in question; to copy the chant text in a Gradual would be more likely to. However, Qui habitat is still assigned to Good Friday in the ninth- to tenth-century Aki5, by which time Eripe me had been regularly appearing in manuscripts for more than a generation. This warns us that the presence of Qui habitat in earlier manuscripts may also reflect local archaisms. The presence of Eripe me in a manuscript securely demonstrates its existence by the time of its compilation, of course, but its absence does not mean that it had not yet been composed nor that it had certainly not yet been encountered by the manuscript compiler.

The chant sources confirm the existence of Eripe me by c. 850. After its earliest dated appearance in Cor2 (c. 853), it appears in Coc6 (860–80), Mon6 (mid-ninth century, or second half of ninth century), Sam2 (ca 875–6), Den7 (877–82), Cha1 (ninth- to tenth century), Gal1 (early-tenth century – before 920), Lan (early-tenth century), Fle1 (tenth century), and, in fact, almost universally in Graduals and Cantatoria from the late-ninth century onwards. Several Ordo Romanus manuscripts from the later-ninth and tenth centuries also attest to its existence at this time.