The increasing use of GIS (Geographic Information Systems) and computational methods in archaeological research runs into various methodological and interpretative roadblocks when used for non-ideal or incomplete datasets. This paper makes use of sampled legacy data on the Indian Lower Palaeolithic obtained by the authors from the Archaeological Survey of India’s annual publication called Indian Archaeology – A Review. It has a three-fold focus: a) How a simple GIS process such as geo-coding can run into various obstacles when the dataset in question is not ideal, i.e. inconsistent in detail and format; b) How preferential archaeological sampling in a region can be better understood through an appreciation of the social, political and economic realities of the time; and c) Additionally, a temporal perspective on the sequence of site surveys, the nature of their reporting and the changing geographic locations of reported sites over two post-Independence decades revealed nuanced information on the changing historical narratives of Indian prehistory within global context. This is a pilot study of the use of geocoding services to enhance and analyse an imperfect and challenging legacy dataset. This study also demonstrates that geocoded sites are not necessarily accurate, but rather indicate the most likely option. The distant structure (ie. top down) and application of GIS make it intrinsically divorced from local reality. Including vernacular sensitivity in GIS, thus making it better suited to work with non-western datasets, would make it far more accurate and useful.