In this thesis, I analyse the textual characteristics of the four Greek, majuscule pandects (01, 02, 03 and 04). By “textual characteristics”, I mean how these manuscripts vary from the initial text as we can best reconstruct it. The term is similar in meaning to “scribal habits”, but rather than referring to the habitual behaviour of the scribe of the manuscript under investigation, “textual characteristics” refers to all the ways in which the manuscript differs from the initial text, both those introduced by its own scribe and by the scribes of all its exemplars. I defend my focus on textual characteristics, rather than scribal habits, by arguing that it is difficult to determine which variant readings were introduced to a manuscript by its own scribe. Royse argues that singular readings were normally introduced by the scribe of the manuscript, but I present a number of arguments against his approach. I also consider the method of Min and Aland, who argue that we should examine textual characteristics by comparing manuscripts to a published critical text. This is likely to be circular, because critical texts are frequently produced using assumptions which favour particular manuscripts. My own method works by attempting to reconstruct the initial text at a range of variation units, giving no weight to “good” manuscripts, just because they are favoured by scholars. I then consider how the manuscripts under investigation differ from this reconstructed initial text. I survey sample chapters in John, Romans, Revelation, Judges and Sirach. The discussion of specific variants in these various books forms the bulk of the thesis. I chose this range of books to be able to survey the wide range of types of literature represented in the pandects and see how textual characteristics varied between them. Working on Judges also allowed me to analyse the “new finds” of 01, which include several pages of text from Judges. In the portion of Judges which I surveyed, 01 and 03 not infrequently agree against all the rest of the B-group of Judges manuscripts, which suggests that they may be closely related. My observations of 04 sometimes challenge the generally accepted transcriptions by Lyon and Tischendorf. In considering the history of the manuscripts, I argue that 01 and 03 may plausibly have been among the manuscripts made in response to the Imperial commissions by Constantine and Constans, recorded by Eusebius and Athanasius. My research has yielded a number of interesting conclusions. In general, across all the manuscripts and for all the varying types of passage sampled, the pandects generally preserve the initial text well. For most pandects, in most books, the mean number of changes from the initial text per ten verses is comfortably below 10.0. Within the changes that can be observed, transcriptional and linguistic variations are more common than harmonisations or changes of content. The more precise profiles of each manuscript vary considerably between Biblical books. The pandects thus create bibliographic unity out of textual diversity. This underlines their significance in the history of the Christian Bible: they reflect in bibliographic form the important hermeneutical move to consider all the books of the Christian Bible as one corpus.