Like many other knowledge workers, professional translators spend much of their day making decisions at a computer: reading texts in one language, searching for background information in many languages, referring to terminology databanks to ensure consistency and appropriacy of lexical choices, checking parallel texts to decide how to meet the needs of the target readership, producing a text in another language, and revising the target text to produce the final translation. If the flow of information or level of concentration is impaired due to non-optimal workplace ergonomics, then the efficiency of the process can be compromised and the quality of the final product can suffer. Despite high levels of competence, professionals may not be able to perform at the expected level or they may have to exert inordinate effort to do so, with potentially detrimental consequences to their health and job satisfaction. In this interdisciplinary research project, models of situated cognition, translation competence, and organizational constraints provide the theoretical framework to consider and evaluate the cognitive and physical ergonomic factors that can impact the situated activity of professional translation. The recursive mixed-methods design combines perspectives from translation studies, occupational health, usability testing, and language technology. It includes ethnographic observation of the workplace, ergonomic assessments, questionnaires, interviews, computer logging, screen recordings, video recordings, eye-tracking, usability experiments, and retrospective verbalizations. A large corpus of data collected in a longitudinal precursor study of translation competence is the source of indications of ergonomic issues. Hypotheses about these are tested and refined with on-site ergonomic assessments and recordings of three groups of professional translators whose work conditions fit various profiles as well as with experiments in a usability lab, an online survey, and in-depth interviews. The data from each project phase are analyzed for indicators of cognitive dissonance attributable to the ergonomics of the human-computer interfaces or workplace conditions and are triangulated with the findings from the other phases. The results of the project are expected to have implications for theoretical models of extended cognition and situated activity, workplace practice, education, and further developments at the human-computer interface as well as to provide insights into how the ergonomics of workplaces can be optimized, not only for professional translators but also for other knowledge workers in similar conditions.