This experiment is part of a study aiming to assess the disturbance experienced by workers in open-plan offices. Previous studies have shown that a sound environment rich with speech sounds can be detrimental to one's performance. The magnitude of this Irrelevant Speech Effect (ISE) depends on the intelligibility of the ambient speech. This has led to the use of the STI to model the induced decrement in performance. However, a decrease in performance is only one aspect of the more general concept of disturbance. When attempting to model the ISE in this regard, other components should be explored. In this first experiment, fifty-six subjects perform a classical seriation task during 10-minutes blocks. They are confronted to sound evironments typical of open-plan offices. In each block, a voice emerges from the noise, with a STI value in the 0.3 – 0.7 range. Both their performance and response times are recorded. A silent condition is used as a reference. After each block, they are presented with the NASA-RTLX questionnaire for a subjective assesment of their workload. Comparisons between speech conditions will be made in order to understand the influence of ambient speech intelligibility on objective and subjective disturbance.