An early printed book seldom informs us directly how it has been produced, but we can distinguish successive phases in the process. For understanding how the text as printed in the book came into being, it is crucial to bear in mind that as long as the book was in production, the structural parts (pages, formes, gatherings) had no relation to the structure of the text as it would be read after its completion. We shall now follow the production process step by step.

1. The printer-publisher had the use of a text, either in manuscript or in print, either owned by or made available to him. This is the printer's copy, or exemplar. If necessary, textual corrections would be entered by a corrector. Before a compositor could begin work, several decisions had to be taken: intended readership, what should the book look like, the format. On this basis the total number of pages was roughly estimated, and following this estimate, the division of pages into quires or gatherings. The next step was to count out in the exemplar the length of each future page, usually by counting lines; the position of the future pages in the gathering was marked in the exemplar as the counting (or 'casting off') proceeded, sometimes by discreet little marks, if the exemplar was valuable or had to be returned to an owner, otherwise by clear letters and figures. Occasionally we can infer that a source was copied especially for use as printer's copy.

2. The compositor placed the prepared exemplar in a clip (the visorium) in front of him. He stood at his type-cases, with lower case and upper case (with the capitals) divided into boxes for each letter, from where he would take up loose type, piece by piece, to form words and lines in his composing stick. After thus setting a few lines he slid them into a galley. In the galley one page was set at a time. The compositor aimed to complete the page at the point indicated in the exemplar. If, however, he did not end precisely at the predetermined point, he might note in the exemplar with his own mark where precisely he had ended the page. He could also adapt the length of his page by adjusting spacing, by variation in spelling and even by adding or omitting a few words. In the early period the compositor's work was by no means purely mechanical; he would use his own understanding of the text and preferences in vocabulary and spelling to adapt it for his contemporary readers.

In the early years the limitations in the supply of type in a printing house made it usually impossible to set a text in the order in which it was to be read ('seriatim'). This would have entailed having at least the first half of a quire plus one page standing in type before printing could begin. Plantin's printing house, specializing in a rich stock of type, was probably an exception to this rule. Composition took therefore place 'per forme'; pages that were combined to be laid on the press and printed together (in a forme) were set in that order, for example, first the outer forme (the first and the last pages) of a quire, followed by the second and one-but-last, etc. This order determined where a compositor absolutely had to end his page as indicated in the exemplar because the following page had already been set and printed.

3. When a page was completed in the galley, it was tied up by wrapping string around it. In this stage a proof would be pulled for the corrector. His task was to mark typographical errors that would then be corrected by the compositor. Typographical errors could be caused by copy being read incorrectly, or the result of 'taking up' mistakes (the compositor had taken a piece of type from the wrong box), or 'distribution errors' (a letter had been returned to the wrong box after it was used). The very rare surviving proofs of the early period show that corrections were indicated by marks that are still in use for this purpose. It could also happen that typographical or other errors were discovered in a later phase during the printing of the forme. When this happened, the correction might be made while the forme was on the press after a number of sheets containing the error had been printed. During printing, type might be loosened by the pressure of the ink ball and be pulled out of the forme, and sometimes typeset matter was damaged by irregular pressure of the platen. In-press corrections and repairs during printing result in copies of a sheet, all belonging to the same edition, existing in variant states. Comparison of copies of the same edition, or comparison with the source that had served as exemplar, may bring us closer to the time in the printing house when the book was made.

4. Once the compositor had completed the required number of pages, they would be imposed to become a forme. The forme consisted of a wooden (later iron) frame in which the pages were firmly fixed with quoins (wooden wedges). The number and combination of pages depended on the format of the book, that is, the number of times the whole sheet was to be folded. In a folio format (one fold) the combination was simple: two pages in a forme, and in quarto (two folds) four pages. The smaller the book, the more often the sheet was folded and the more complicated the imposition scheme. Printers' manuals show imposition schemes in all their varieties. Many early editions still bear the traces of imposition errors.

5. The imposed forme was now laid on the press. Early printing presses were built of wood and consisted of a stone or marble bed, a platen, a movable carriage that made it possible to ink the typeset forme and place a paper sheet over it for each impression by folding over the tympan and the frisket that were attached by a hinge. The movable carriage was first used around 1472 in Rome, and was introduced in the Netherlands around 1480. This technical improvement made it possible to print a whole sheet with two pulls of the press. Before then a simpler press was used, on which only half a sheet could be printed at one time. This meant that formes had only half the number of pages compared to those printed on presses with movable carriages. For quartos, divided half-sheets were used.

6. After the forme was solidly secured on the bed of the press, it was evenly inked before each impression with two large ink-balls filled with horse hair that were rubbed on an ink block. The sheets of paper (or vellum) were slightly dampened in advance. Each sheet to be printed was laid on the tympan where press points ensured that it was positioned accurately. A light hinged frame (the frisket) was closed over the sheet on the tympan, and the whole was flipped over onto the inked forme. The carriage slid the forme below the platen. The platen was lowered using a horizontal lever (the boom) that was connected to the spindle supporting the platen, and by its contact with the inked form the impression was made. The whole print run, sheet after sheet, was printed in this way on one side. The sheets were hung up to dry a little, after which the procedure was repeated for the other side of the sheets, the backing up. The points ensured that the pages matched precisely ('kept register') during the printing of the first forme ('the white paper') and the backing.

Gutenberg invented not only movable type but also printing ink. Ink used for writing was water-based, and could not adhere to metal type. Printing ink was composed of oil-based varnishes and pigments including soot for black ink and mercury for red. Ink-making was a specialised skill, and printers found it often difficult to obtain ink of good quality.

7. After printing was completed, the sheets were hung out to allow the paper to dry. There is some evidence that sheets were usually stored as they had come from the press, in any case until the printing of the entire book was completed. Not until they were sold were the sheets gathered to form copies of the book. There are, however, exceptions.

8. After the print-run of each forme was completed, it was taken off the press and washed with lye in order to remove all traces of ink. The pages were then loosened and the type distributed, that is, returned to the appropriate boxes of the type-cases. Sometimes mistakes were made, which led to typographical errors when the type was used again. If more than one fount of type was used mistakes could be made in keeping the founts apart, with visible results in subsequent printing.