The aim of this paper is to show that the manipulation of a piece of equipment can affect the revision of its accompanying procedural document. An experiment was set up to investigate how the presence or absence of the equipment might influence the elaboration of representations at different levels (the so-called verbatim, semantic and pragmatic representations). Results show that, when writers do not have direct access to the equipment, they focus mainly on the elaboration of verbatim representations when revising the text. By contrast, when given the opportunity to manipulate the equipment, writers mainly focus on the elaboration of pragmatic representations. In the former case, the proportion of revisions made to surface aspects of the text (such as spelling and syntax) was higher than the proportion of revisions made to the pragmatic aspects (such as qualifiers of the actions to carry out and indexical processes). In the latter case, the proportion of revisions made to pragmatic aspects of the text was higher than those made to surface aspects of the text. Moreover, correlation analyses show a negative correlation between the proportion of changes made at the verbatim level and the proportion of changes made at the pragmatic level. It also reveals a negative correlation between the proportion of changes made at the verbatim level and the proportion of changes made at the semantic level. In order to complete the experiment, three different documents were evaluated in a comparative study: the manufacturer?s original instructions, instructions revised with the equipment, and instructions revised without the equipment. Results showed that the decreasing order of preference chosen by the judges was as follows: (1) the instructions revised while manipulating the equipment; (2) the instructions revised without the equipment and (3) the manufacturer?s original instructions. The paper concludes by considering the theoretical and practical aspects gained from this research. From a theoretical point of view, it provides evidence of the elaboration of different levels of representation whilst revising the procedural instructions. From a practical point of view, it relies on the different levels of representation to give recommendations for revising procedural documents. For example, the revision of procedural documents could rely on the combination of various approaches, such as adopting a revision process in the absence of the equipment and a revision process while manipulating the equipment.