It is common practice nowadays to assess the presence of terrigenous (land-derived) sediments in deep-sea cores using bulk geochemical data, but the key issue is to identify the source of these sediments and the way they were transported to the core site in order to interpret their palaeoclimatic significance. Here, we demonstrate a new approach taken to geochemically-fingerprint a large set of sediments collected from potential source areas (PSAs) in southeastern and southcentral Australia and to compare these data with the record obtained from X-ray Fluorescence (XRF) scanning on a long deep-sea sediment core MD03-2607 obtained offshore Kangaroo Island, South Australia. The entire data set of samples collected on land as well as the downcore measurements were unmixed using the numerical end-member method AnalySize. We successfully use the elements Al, Fe, K, Mn, S, Sr and Y to define end members. In addition, the on-land occurrences of the chemical ratios of Zr/Zn, Ti/Rb, Ti/Y and Zr/Rb are used to support the provenance of the chemical end-members. Three main PSA's are defined: Murray River Basin (MRB), Darling River Basin (DRB) and Kati Thanda – Lake Eyre District (LED), of which the MRB is represented in two different chemical end members. The downcore contributions of these end members in the sediment core are consequently interpreted in terms of fluvial (MRB and DRB) versus aeolian (LED) input. We determined the origin of the terrigenous sediments recovered from the core for the last glacial-interglacial cycle that also has implications for atmospheric circulation across southern Australia.