TNO and BMA Ergonomics are developing a so-called smart office chair. This chair is supposed to provide feedback on postures and movements during seated office work. The feedback should enable the user (i.e. the worker doing VDU tasks) to perform his or her work with less discomfort and in a more productive way. One part of this project is a study about the development of musculoskeletal discomfort during VDU work, sitting in various postures. Nineteen subjects did 2 types of VDU work in a laboratory setting: call-centre and data-entry. They were instructed to work in a certain sitting posture for 30 minutes, without pauses. After that, they could relax for 30 minutes. Four sitting postures were imposed; the 5th one was each individual’s freely chosen working posture: the reference posture. Every 3 minutes, local perceived discomfort (LPD) of 5 body regions was determined, by letting subjects rate their LPD on a 10-point-scale. From that we calculated the relative LPD-dose, setting the dose of the ergonomic optimal posture to 100%. LPD-doses were not significantly different between the 2 VDU-tasks. Different sitting postures did lead to significantly different LPD-doses; the lowest in the ergonomic optimal posture and the highest one in the ‘vulture’ posture. From the individual development of LPD during 30 minutes of VDU work and a prior evaluation of the health risks of prolonged LPD, we can deduce the period of time a subject is allowed to work in a specific posture. These data can serve as input for a demonstration model of the smart office chair.