We employed a novel approach using the bipolar and tropical distributions of bivalves to track the northward drift of the Indian Plate. The Kimmeridgian-Tithonian Retroceramus, Tithonian Anopaea and Buchia, and Aptian-Cenomanian Aucellina were mainly restricted to areas poleward of 30 degrees N and 30 degrees S, but middle Oxfordian-latest Cretaceous rudist bivalves were mainly restricted to areas between the paleolatitudes 30 degrees S and 30 degrees N. The distributions of these paleolatitude indicators in late Mesozoic time record five major drifting phases of the northward translation of the Indian Plate. In the Late Jurassic-earliest Cretaceous, the Indian Plate was a part of Gondwana, and various bipolar bivalves were present. During the Aptian-Albian interval, the Indian Plate was separated from Gondwana, but most parts of this plate were still under cold water, likely south of 30 degrees S, and even at higher latitudes, because bipolar bivalves still existed at the northern margin of this plate, in the Tethyan Himalaya. In the Cenomanian, the northern part of the Indian Plate translated north into a warm/subtropical environments, north of 30 degrees S, but southern India was still in a cold region, south of 30 degrees S, as evidenced by Aucellina, which had already disappeared from the northern part of the Indian Plate, but still survived in southern India. In the latest Campanian, at least the Tethyan Himalaya of the northern margin of the Indian Plate had drifted into the area north of 30 degrees S, based on the appearance of Bournonia haydeni then in Tibet. Rudists occupied both the northern (Tethyan Himalaya) and southern (southern India) parts of the India Plate by the Maastrichtian, which demonstrate that the Indian Plate completely entered the tropical area, and roughly coincided with the initial collision between the Indian and Eurasian continents around the Maastrichtian-Paleocene.