The purpose of the article is to describe visual context as one of the possible means of teaching technical translation. Using multimodal visual contexts, a student of technical translation is able to assimilate technical knowledge at the stage of pre-translation, which leads to understanding the content of a technical text and, as a result, to high-quality conscious translation. The work has stated that visual material is more efficiently classified in terms of its most important functions and in terms of its connection with the real world that it represents. Therefore, I proposed to take into account the following characteristics of the visual material: visual or verbal ways of presenting knowledge should be related in time and space; when choosing visual materials, excessive information should be avoided; dynamic phenomena are better represented through animation; highly interactive elements should be presented separately for beginners. Among these characteristics there are two main ones-iconicity and dynamism. The results of a comparative analysis of technical texts about the oil and gas industry showed that contexts are differentiated, taking into account two main characteristics of images, into the following types: static verbal-visual contexts with a low level of knowledge, static visual contexts with a low level of knowledge, dynamic verbal-visual contexts with a high level of knowledge, and dynamic visual contexts with a high level of knowledge.