Although the importance of peat as energy supplier in the medieval and early modern North Sea Area is well known, the location, extent and nature of the peat-producing areas—peat marshes or mires—remains amongst the major problems in the landscape history of the coastal wetlands. This is especially true for areas like Northern Flanders, where peat marshes have since completely disappeared. This article reconsiders the ‘peat debate’ between geoscientists, who rely on ‘positive’ soil evidence, and historians, who accept ‘circumstantial’ historical data so as to reconstruct former peat marshes. Based on a systematic comparison of the arguments of both geoscientists and historians, we argue that recent methodological advances, such as integration of historical and geophysical data in a GIS, allow for bridging the gap between the two approaches and to reconcile contrasting opinions on historical peat marshes. This is tested in a case study for two villages—Moerbeke and Wachtebeke—in Northern Flanders, where re-evaluation of both geophysical features (soil, elevation models, hydrology) and archaeological and historical data (maps, documentary evidence on landed property and peat extraction and toponyms) leads to a completely new model for the presence of (Holocene) peat marshes.