Archaeological research was undertaken in 2011/12 on the isolated island of Rossel in the Louisiade Archipelago, Milne Bay Province, PNG. The small population living on the island today speak a language that is unrelated to the languages spoken elsewhere in this vast island region. The complexities of their cultural traditions are also regionally unparalleled, and the people themselves are genetically distinct. Rossel Island is therefore considered to be an enigmatic outlier in an otherwise culturally similar region. Previous archaeological research on other islands in the Milne Bay province indicates that people have occupied the region for around 2800-2600 years. The first settlers in the region were pottery producing communities, possibly associated with the widespread Lapita Cultural Complex. However, linguistic and anthropological theories attempting to explain the social complexity on Rossel Island suggest that the island may have been colonised by a different cultural group altogether. How and when this might have occurred has been of great interest to early explorers and modern researchers alike. The aims of the Rossel Island project were therefore twofold. Firstly, to develop a chronological model for human occupation on Rossel Island, and secondly, to determine if spatial and/or temporal changes in the archaeological record may have been influential towards the development of the unique Rossel Island culture. To place Rossel Island in a comparative regional framework, excavation was also undertaken on Nimowa Island, 80 km to the west. This paper will present the AMS dates, as well as the preliminary results from the material culture analyses. An understanding of prehistoric cultural development on Rossel Island and in the Louisiade Archipelago will then be developed.