The localization of the city of Liège, in the Meuse (Maas) valley, later the development of its industries and its port, were all conditioned by the natural geography of its site. Yet today the site itself and its natural characteristics are seldom perceptible from within the city. The large scale destructions in the city fabric during the last two centuries broke up its historical siteresponsive urban morphology. The disappearing industry left the landscape scarred with now meaningless traces, and no strong alternative vision for its future. While the region is waiting for “the” solution, for the definitive project that will give it new impulse and identity, the abandoned industrial sites and many of their surroundings are turning into “non-sites”. To gain a new vision specific to the valley, one must read todays’ landscape as a palimpsest left by rural uses and urban developments, including industrial artifacts, overlaid on its original geomorphology. A “project” approach, such as that proposed by many landscape architects at different scales, from garden to forest and larger urban development, offers a way of observing and interpreting the landscape, eventually leading to iterative, local interventions, (“landscape acupuncture”). Inspired by the site-responsive agricultural past of the valley, the interventions should lead to a new spatial language of urban agriculture and forestry. Each of these interventions can turn into a “landscape laboratory”, involving local stakeholders, whose aim is to articulate small scale landscape elements with long-term place and time-contextual investment, and thus redefine its identity. ; Peer Reviewed