Studying the erosion record of a mountain belt is useful for understanding processes of mountain building, such as crustal deformation and the coupling of tectonics and erosion. It is also important for understanding its potential affect on regional and global climate and changes to marine geochemistry. With respect to the Himalaya, erosion has been suggested as a primary cause of global cooling and changes to ocean geochemistry in the Cenozoic. Often the erosion record preserved in a mountain's sedimentary repositories are the only way of gaining information of these processes in the absence of a hinterland record, which may have been overprinted by metamorphism or eroded from the mountain belt itself over time. Little is known of the early erosive history of the Himalaya. In the foreland basin a widespread unconformity separates pre-collisional sedimentary rocks from Himalayan-derived sedimentary rocks. Furthermore, the record of collision and erosion preserved in the suture zone does not relate to erosion of the southern flank and the potential record stored in offshore fans such as the Indus and Bengal is on the whole poorly constrained due to current drill depth restrictions. However, it has been proposed that an early record of Himalayan erosion may be preserved in sedimentary deposits found in the Andaman Islands and the Indo-Burman Ranges/Chittagong Hill Tracts as offscraped material from the proto Bengal Fan. My key aim is to determine the provenance of the Tertiary sedimentary (accretionary prism) rocks that run along the India-Asia (Sunda) subduction zone from Bangladesh and Burma (Myanmar) to the Andaman Islands. To achieve this a unique multi-proxy approach is used. Three main conclusions can be drawn. Firstly, that the Palaeogene sedimentary rocks of both South Andaman and the Indo-Burman Ranges (Burma) show appreciable magmatic arc derivation with additional orogenic input. Although a subordinate continental source to these regions is observable from the data it is most likely that this source is the Burman continental margin and not the Himalaya. However, I do not altogether rule out the possibility of minor Himalayan contribution. Unambiguous evidence of Himalayan derivation is not found in the region until ~38 Ma in the Bengal Basin. In younger Palaeogene to Neogene sedimentary rocks of the Andaman Islands (30-20 Ma) a provenance change to dominantly recycled orogen with minor arc sources is seen. This is most easily explained by derivation from Burman continental margin sources and not from the Himalaya. Secondly, the Neogene rocks west of the Kaladan Fault, preserve evidence for Himalayan derivation which is unambiguous. The rocks in the westernmost Indo-Burman Ranges and the Chittagong Hill Tracts/Hatia Trough are isotopically and petrographically distinct from their Palaeogene counterparts, but similar to coeval rocks of known Himalayan derivation in the orogen's foreland basin. Thirdly, seismic mapping of the Neogene sedimentary rocks have given new insight to palaeodrainage of the Brahmaputra. New seismic data suggests that it was not the uplift of the Shillong Plateau at ~3. 5 Ma that caused the deflection of the Brahmaputra around the northern flank of the massif, but instead the encroachment of the leading edge of the growing Chittagong Hill Tracts onto the already uplifted plateau that closed the northeast drainage route to the Brahmaputra, forcing it to divert its course. Our data is consistent with the hypothesis that erosion of the Himalaya could have caused an increase in global ocean 87Sr/86Sr ratios since ~40 Ma (Richter et al. 1992). However, it does not appear to be consistent with the theory that onset of substantial erosion from the Himalaya at ~55 Ma caused global cooling, due to the lack of any substantial evidence as yet, of erosion since collision. The results of this thesis support models of crustal deformation, which require a delay in erosion or do not require erosion at all in early stages of convergence.