The merging of two intensive care units is a time of profound change, and constitutes a risk of mishaps. We report some aspects of such a project in our institution. The evaluation of various indicators reflecting the activity, patient's hospital pathways, mortality, as well as the use of specific techniques, has shown that no particular problem was observed during the first 9 months. Improvements in performance or productivity have not been demonstrated so far. The follow-up will permit to demonstrate long-term benefits. We believe that these observations may be of interest for other departmental or hospital reorganisations.