The distribution of settlements in ancient societies is commonly linked to the courses of large river systems. The Bronze Age Harappan civilisation (4800-3500BP) is no exception with the major sites of Harappa and Mohenjo-Daro located along the Indus river system. However, the largest collection of Harappan archaeological sites are associated with the postulated surface trace of a large palaeo-river channel in the drainage divide tract between the Ganges and Indus river systems, where no major river currently flows. It has been proposed that this palaeo-channel was occupied by a major river sourced in the Himalaya, and that this river provided water resources to sustain the extensive Harappan sites located along its ancient course. The abrupt abandonment of urban centres here at ~3500 BP has been explained by explanations such as river diversions and abrupt monsoon weakening. These hypotheses have remained untested because the stratigraphy and chronology of the postulated palaeochannel has never been determined. Here we present an integrated study of the Ghaggar-Hakra palaeochannel based on detailed analysis in NW India. We use a combination of satellite image analysis, subsurface geophysical analysis and detailed sediment coring to analyse the large-scale planform geometry, and detailed sedimentary and stratigraphic nature of the postulated palaeochannel in NW India. In particular we focus our analysis on a tract of the proposed channel adjacent to the major Harappan urban centre of Kalibangan in Rajastan State. We will consider how changes in the course of the palaeochannel may have influenced settlement patterns of the Harappan civilisation in present day NW India.?