Pollen records from sediment cores do not allow understanding of past human use of Cannabis because Cannabis may also grow naturally or increase incidentally in the presence of human settlements. This was particularly true in Asia, where Cannabis was not used as hemp as often as it was in Europe; hemp retting makes it easy to identify Cannabis use as it causes characteristically high increases in Cannabis fossil pollen in strata where it was used. In this study, we evaluate the evidence for Cannabis cultivation and use in Central India, where Cannabis fossil pollen occur in sediment cores from ~ 12.2 ka (1 ka= 1000y ago). While the public perception is that Cannabis has a long history of use in South Asia, textual sources suggest significant increases in Cannabis use ~1-2 ka. To disentangle Cannabis presence due to natural dispersal and due to use, we develop a new approach. Artemisia grows alongside Cannabis and disperses in similar conditions, whether naturally or due to increase in human settlements. We investigate when Cannabis increased in the paleo-record independent of Artemisia and find that Cannabis increased ~ 2.5 ka in areas close to major early historic settlements, where Cannabis displays characteristic patterns of cultivation. Cannabis does not display these patterns away from the early historic settlements. These dates are similar to findings of Cannabis use as hemp in the Himalayas (2.5 ka) and as a drug in China (2.7 ka), but pre-date the proliferation of Cannabis presence in textual sources. Overall, Cannabis pollen presence in pre-historic India (<2.5 ka), although common, is likely to be because of natural dispersal of Cannabis.