The source of the gold (Au) in the Au-bearing stream sediments within the central parts of Tanzania is poorly constrained. Hence, this study aims to elucidate the sources of these Au-bearing sediments and gold using whole-rock geochemistry of stream sediments and multivariate statistics. From the CIA values, the sediments are weakly weathered and compositionally immature, considering the ICV and Fe/K values, under humid climatic conditions. The source ratio proxies: Th/Sc, Th/U, Cr/Th, La/Sc, and Th/Cr, indicate a felsic igneous rock source. The discriminant plots show intermediate and mafic rocks, while the multivariate indicates a dominant mafic source rock with 40.35% relative to 18.70% extraction variance for felsic rocks. The Au, however, does not associate with mafic rocks (Ni, Co, V, and Cr) or felsic rocks (Th, Sr, Rb, and K), with all these independent parameters revealing a > 0.05 p-value from the multilinear regression model (MLM). The study will recommend the use of a multi-method approach in the search for Au within the central parts of Tanzania, where soil/rock geochemistry is combined with stream sediment geochemistry and the data processed using multivariate statistics together with machine learning methods, e.g., artificial neural network.